Tag: cosmetic claims

Firmness Actives in Korea: Wrinkle Is Notified, Elasticity Is Substantiated

Hello, this is Dr. J. Today, five actives people reach for when they want firmer skin — chosen by what the human studies actually measured.

Start with the odd part. In Korea, the regulator reviews “wrinkle improvement” and names the actives for it. It does not use the word elasticity anywhere in that table. Instead it wrote down how to measure elasticity, in a different document.

――――――――――――――――――――――――――――

Primary sources

📄 Functional Cosmetics Review Regulation, Annex 4 (Korea Law Service)

📄 MFDS, Guideline on test methods for claim substantiation (March 2018, PDF, Korean)

📄 Hyaluronic acid, 8-week study (J Clin Aesthet Dermatol 2014)

📄 0.4% retinol, 24-week trial (Arch Dermatol 2007)

Annex 4 holds six claim categories. The guideline holds ten test methods. Elasticity appears in the second document only.

――――――――――――――――――――――――――――

The short version

● Korea notifies six claim categories, and wrinkle improvement names four actives
● The word elasticity is not in that table
● Chapter 2 of the substantiation guideline is “Skin elasticity improvement”
● A Cutometer® or Dermaflex® applies negative pressure and measures the return
● So the test to apply to a product is: what did the study measure?

――――――――――――――――――――――――――――

How Korea splits one face into two boxes

Korean cosmetics law carves out a class called functional cosmetics. For each claim in that class the regulator publishes an annex that names the active and the percentage, and a product built to those numbers is spared from filing new efficacy data. Annex 4 is that table.

Category	What the annex sets	Reading
UV protection and tanning	Actives and maximum percentages	Both the name and the number are fixed
Whitening	Nine actives (arbutin 2–5%, niacinamide 2–5%, and others)	Same
Wrinkle improvement	Four actives (retinol 2,500 IU/g, adenosine 0.04%, and others)	The box next to today’s subject
Hair colour change	A list of dye actives	Same
Hair removal	Thioglycolic acid 3.0–4.5%	A single active
Acne relief	Salicylic acid 0.5%	Same
Firmness / elasticity	No category at all	Not a claim the regulator reviews

The wrinkle box, in full

Active	Percentage	Reading
Retinol	2,500 IU/g	One of the few set in international units
Retinyl palmitate	10,000 IU/g	A larger number than retinol
Adenosine	0.04%	The smallest percentage of the four
Polyethoxylated retinamide	0.05–0.2%	Written as a range

Annex 4, wrinkle entry — as written

Dosage forms: “limited to lotions, solutions, creams and impregnated masks”
Claim: “helps improve skin wrinkles”
Directions: “apply an appropriate amount evenly to the skin; or apply the product to the skin, remove the support after 10–20 minutes and spread the remaining product evenly (impregnated masks only)”

Even the sentence a brand may print is fixed. Elasticity is in none of those cells.

――――――――――――――――――――――――――――

What exists instead: a written test method

The second document is a civil petitioner guideline published by the Ministry of Food and Drug Safety in March 2018. It sets out how to substantiate ten claims that fall outside the functional class.

Chapter	Claim	Added
Ch. 1	Skin moisturisation	2012.5 (original)
Ch. 2	Skin elasticity improvement	2012.5 (original)
Ch. 3	Sebum control	2012.5 (original)
Ch. 4	Temporary cellulite reduction	2013.11
Ch. 5	Suitable for acne-prone skin	2013.11
Ch. 6	Dark circle relief	2014.8
Ch. 7	Improved skin blood flow	2015.12
Ch. 8	Puffiness relief	2015.12
Ch. 9	Paraben-free	2017.3.31
Ch. 10	Fine dust protection	2018.3.28

Chapter 2, Skin elasticity improvement — the guideline’s own sentences

“This is a test in which a Cutometer® or Dermaflex® applies negative pressure to the skin and measures the degree to which it returns to its original state”
“The number of subjects shall provide at least 20 sets of valid data so that statistical comparison is possible”
“from among adult men and women aged 20 or over”
“The test site shall be the whole face, including the eye area (the measurement site)”

The conditions are written down as well: no air movement or direct sunlight, 20–24 °C and 40–60% relative humidity, 30 minutes of acclimatisation before measurement, a significance level of 0.05, and non-parametric tests as the default when an arm holds 30 subjects or fewer.

⚠ What Chapter 2 does not contain is as telling. There is no list of permitted wordings and no pass mark, and no parameter is named — R2, R5 and R7 appear nowhere. The state writes the procedure and leaves the threshold open.

――――――――――――――――――――――――――――

One study measured elasticity directly — hyaluronic acid

33 women, mean age 45.2 ± 3.9, over 8 weeks at a contract institute in Hamburg. A low molecular weight hyaluronic acid lotion, serum and cream were applied twice a day to the whole face, with measurements at baseline and on days 15, 29 and 57.

Elasticity was read with a Cutometer, and the treated side rose significantly above the untreated side. Moisture, roughness and erythema were measured too; the only p value printed in the body of the paper belongs to erythema.

The weak points belong in the same paragraph. The “up to 55%” figure lives in the abstract and is not carried into the text, no p value is given for elasticity, and there is no placebo arm — the comparison is with an untreated area. The concentration of the product is not disclosed either.

――――――――――――――――――――――――――――

Two are notified actives — retinol and adenosine

The retinol trial measured wrinkles. 36 residents of two long-term care facilities, mean age 87, were followed for 24 weeks in a randomised, double-blind, vehicle-controlled design: 0.4% retinol lotion on one arm, its vehicle on the other, up to three times a week.

Fine wrinkling scores changed by −1.64 (95% CI −2.06 to −1.22) with retinol against −0.08 (−0.17 to 0.01) with vehicle, P<.001. In biopsy subsets, glycosaminoglycan expression rose (P=.02) and procollagen I staining increased (P=.049).

That trial measured no elasticity. It measured wrinkle scores and tissue markers. Adenosine, meanwhile, carries a number the regulator chose: 0.04%, set in Annex 4, not by the brand.

――――――――――――――――――――――――――――

One looked at the fibres — vitamin C 5%

A 5% ascorbic acid cream was compared with its own excipient for 6 months, double-blind, applied daily, with assessments at baseline, 3 months and 6 months and biopsies at the end.

Global scores from both the dermatologist and the volunteer improved against control, the density of the skin microrelief rose and deep furrows fell. Electron microscopy showed repair of elastic tissue, consistent with the clinical and surface findings.

Two limits: the number of participants is not stated in the abstract (healthy female volunteers with photoaged skin is as far as it goes), and what this study saw is tissue, not an elasticity reading.

――――――――――――――――――――――――――――

One was measured by mouth — collagen peptides

69 women aged 35–55, 23 per arm, took 2.5 g or 5.0 g of collagen hydrolysate or placebo once daily for 8 weeks, with a 4-week follow-up after the supplement stopped.

Both doses improved skin elasticity against placebo at 8 weeks, and four weeks after stopping, elasticity remained significantly higher in the older women. Hydration and transepidermal water loss moved in subgroup analysis but not significantly.

This one is not applied to the skin. It belongs on a different shelf from the jar, and the exact figures, p values and the instrument used sit behind a paywall this piece could not open.

――――――――――――――――――――――――――――

The five in one table

Active	Subjects	Duration	Design	Measured	Result
Hyaluronic acid	33	8 weeks	Treated vs untreated area	Elasticity	Significantly higher on the treated side
Retinol 0.4%	36	24 weeks	Randomised, double-blind, vehicle	Wrinkles, tissue	−1.64 vs −0.08 (P<.001)
Adenosine 0.04%	—	—	Set by Annex 4	—	Notified for wrinkle improvement
Vitamin C 5%	not in abstract	6 months	Double-blind, vehicle	Elastic tissue	Repair on electron microscopy
Collagen peptides	69	8 weeks	Randomised, double-blind, placebo	Elasticity (oral)	Both doses above placebo

And where each one is weak

Active	What this study cannot say
Hyaluronic acid	The 55% figure is abstract-only; no p value for elasticity and no placebo arm; the concentration is undisclosed
Retinol	It did not measure elasticity. The cohort is also a mean age of 87 in residential care, which does not transfer cleanly to younger skin
Adenosine	What the regulator approved is wrinkle improvement, not an elasticity figure
Vitamin C	No participant count in the abstract, and it looked at tissue rather than an elasticity reading
Collagen peptides	Not a topical result. The numbers, p values and instrument could not be verified in the full text

――――――――――――――――――――――――――――

Korea, the United States and the EU — reviewed names and proven names

	Korea	United States	EU
Ingredient list per claim	Annex 4 of the functional cosmetics regulation — actives and percentages for six claims	None	Positive lists cover colorants, preservatives and UV filters only (Regulation 1223/2009, Art. 14)
“Wrinkle”	Reviewed or notified before sale as a functional claim	No approved list of claims	Handled as a claim, under the common criteria
“Elasticity”	Chapter 2 writes down the measuring procedure	No procedure written by the regulator	The six common criteria demand evidence
In their words	“a Cutometer® or Dermaflex® applies negative pressure to the skin…”	“FDA does not have a list of approved or accepted claims for cosmetics”	“adequate and verifiable evidence” · “state of the art practices” (655/2013)
When it is checked	Functional claims before sale; elasticity when asked	No pre-market approval for cosmetics	No pre-market approval; the evidence must exist

Lined up, the Korean position is unusual in a specific way. One face, two boxes. The claim closest to firmness — wrinkles — is handled by ingredient, in advance. Firmness itself is handled by procedure, afterwards. Neither the United States nor the EU keeps the first box for this kind of claim at all, which is why an exporter reads the two sentences in opposite order: in Seoul the first question is which active and what percentage, and in Brussels or Washington it is what evidence is on file.

――――――――――――――――――――――――――――

What was left out, and why

✔ The niacinamide elasticity trial (Bissett et al., Dermatol Surg 2005) — the abstract could not be opened at a primary source today. Europe PMC returned 429 on all three attempts and Wiley, the journal’s own full text and Semantic Scholar all errored
✔ Injected hyaluronic acid (mesotherapy) — a procedure, not a cosmetic
✔ Lasers, radiofrequency and thread lifts — same reason
✔ Madecassoside — the primary sources found were reviews
✔ A green tea study (Chiu 2005) — oral and topical treatment were given together, so neither arm can be read alone
✔ A pass mark for Chapter 2 — the guideline does not set one

About the author · Dr. J

MS in Veterinary Medicine, PhD in Life Sciences. I worked as a researcher at the Korea Research Institute of Bioscience and Biotechnology (KRIBB) and now run Beaurea, a cosmetics manufacturing company, where I write the formulas myself. This blog reads the research papers and the regulatory documents and turns them into plain-language criteria for choosing ingredients. I don’t recommend specific products here – only criteria that apply to any product.

📌 Related

📌 Adenosine 0.04%: who decided that number — the box next door, wrinkle improvement

📌 Hyaluronic acid: how to read the molecular weight — the one active here that measured elasticity

📌 Collagen: topical vs oral — the longer version of the oral section

📌 Pores and sebum: the test method the regulator wrote down — Chapter 3 of the same guideline

📄 References

· Functional Cosmetics Review Regulation, Annex 4 (Korea Law Service, read 21 October 2026)

· MFDS, Guideline on test methods for the substantiation of cosmetic labelling and advertising (civil petitioner guideline, March 2018), Chapter 2

· Jegasothy SM, Zabolotniaia V, Bielfeldt S. J Clin Aesthet Dermatol 2014;7(3):27-29

· Kafi R, Kwak HSR, Schumacher WE, Cho S, Hanft VN, Hamilton TA, King AL, Neal JD, Varani J, Fisher GJ, Voorhees JJ, Kang S. Arch Dermatol 2007;143(5):606-612

· Humbert PG, Haftek M, Creidi P, Lapière C, Nusgens B, Richard A, Schmitt D, Rougier A, Zahouani H. Exp Dermatol 2003;12(3):237-244

· Proksch E, Segger D, Degwert J, Schunck M, Zague V, Oesser S. Skin Pharmacol Physiol 2014;27(1):47-55

· U.S. FDA, “Cosmetics Labeling Claims” (content current as of 21 November 2022)

· Commission Regulation (EU) No 655/2013, common criteria · Regulation (EC) No 1223/2009, Article 14

――――――――――――――――――――――――――――

This article summarises published research and public regulatory documents; it is not medical advice. Annex 4 and Chapter 2 of the substantiation guideline were read at source on 21 October 2026, and the Korean documents are quoted here in translation. Of the four studies, one does not state its participant count, one measured no elasticity at all, and one was taken by mouth. Those facts are in the tables. Ingredient names and percentages are given as criteria for reading a product, not as a recommendation of any product.

— Dr. J Ingredient Notes

October 21, 2026

Exfoliating Actives in Korea: No Claim Category, but a Label Rule

🇰🇷 한국어판

Hello, this is Dr. J. Today we are looking at exfoliating actives in Korean cosmetics, and at what the studies behind them actually measured.

Here is the part most guides skip. Exfoliation is not a reviewed claim in Korea. It is not one of the six notified categories, and it is not one of the ten official test methods. Yet the label rules name it anyway.

――――――――――――――――――――――――――――

The primary documents

📄 Regulation on the Examination of Functional Cosmetics, Annex 4

📄 Regulation on Safety Standards for Cosmetics, Annex 2

📄 MFDS Guideline on Test Methods for Substantiating Cosmetic Claims (March 2018, PDF)

📄 U.S. FDA, Alpha Hydroxy Acids

📄 Stiller MJ et al., Arch Dermatol 1996 (22 weeks, 74 women)

Annex 4 names six categories; the guideline names ten test methods. Exfoliation is in neither.

――――――――――――――――――――――――――――

The short version

✔ Korea notifies actives in six categories, and exfoliation is not one of them
✔ The regulator’s ten test methods do not cover it either
✔ Above 0.5% AHA a sun sensitivity warning becomes mandatory on the box
✔ Even a 10-50 mL package that shortens its ingredient list cannot drop fruit acid (AHA)
✔ So the useful question is what each study measured

――――――――――――――――――――――――――――

1. Korea notifies actives by claim, and exfoliation is missing

Korea runs functional cosmetics as a positive list. For each claim the notification names the ingredient and the percentage, and it also fixes the dosage form and the wording a company may print.

Category	What the notification sets	How to read it
UV protection and tanning	27 ingredients with maximum levels	Both the name and the number are fixed
Brightening	Nine ingredients (arbutin 2-5%, niacinamide 2-5% and others)	Same
Wrinkle improvement	Four ingredients (retinol 2,500 IU/g, adenosine 0.04% and others)	Same
Hair colour change	Dye lists split across six tables	Same
Hair removal	Thioglycolic acid 3.0-4.5%, pH 7.0 to below 12.7	A single ingredient
Acne-prone skin relief	Salicylic acid 0.5%, in cleansing products of soap composition	One of today’s five sits here
Exfoliation	There is no category	So there is nothing to be reviewed for

The last row is the point of this article. There is no exfoliation category, so there is nothing for a company to be reviewed against, and no notified percentage to compare a product with.

――――――――――――――――――――――――――――

2. There is no test method for it either

Separately from the notification, the Ministry publishes a guideline that sets out how to measure each claim: who the subjects are, which instrument is used, and what counts as a result. It has grown one claim at a time.

Chapter	Claim	Added
1	Skin hydration	2012, at the start
2	Skin elasticity	2012, at the start
3	Sebum control	2012, at the start
4	Temporary reduction of cellulite	2013
5	Suitable for acne-prone skin	2013
6	Dark circles	2014
7	Skin microcirculation	2015
8	Puffiness	2015
9	Paraben-free	31 March 2017
10	Fine dust protection	28 March 2018
—	Exfoliation	Never added

Ten chapters, added over six years, and exfoliation never arrived. In our earlier pieces the pattern was different each time: for sensitive skin the phrase appeared only as a subject exclusion criterion, and for sebum there was a full chapter with a Sebumeter procedure. Exfoliation is the third case — absent from both documents.

――――――――――――――――――――――――――――

3. The label rules name it even so

A claim category is one instrument. A label rule is another, and Korea uses the second here. Once a product passes 0.5% AHA, a notification requires fixed sentences on the box.

When it applies	What the box has to say
AHA above 0.5%	The product may increase the skin’s sensitivity to sunlight, so use a sunscreen alongside it (rinse-off and hair products are excepted).
Same threshold	Test it on a small area first and check for a skin reaction.
AHA above 10%, or pH below 3.5	The product contains a high concentration of AHA and may cause adverse effects, so consult a physician.

So the sunscreen sentence you find on the back of a Korean exfoliating toner is not marketing. It is text a notification requires, and its trigger is a concentration.

The second place is the ingredient list. Small containers may shorten it, but six things stay.

Enforcement Rule of the Cosmetics Act, Article 19(2)3 — what a small package may not drop

A package holding more than 10 mL and up to 50 mL (or more than 10 g and up to 50 g) may omit ingredient labelling, except for:
(a) tar colourants · (b) gold leaf · (c) the type of phosphate in shampoos and rinses
(d) fruit acid (AHA) · (e) ingredients that produce the effect of a functional cosmetic · (f) ingredients for which the Minister has notified a blending limit
※ Where labelling is omitted, the company must let consumers check the full list by telephone or website, or keep a printed list at the point of sale (paragraph 5).

An active with no claim category of its own is still named in the labelling rules. That is the shape of this subject in Korea.

――――――――――――――――――――――――――――

4. Roughness: the two AHAs did not behave the same

The cleanest cosmetic-strength trial here is old and still useful. 74 women aged 40 to 70, 22 weeks, randomised, double blind and vehicle controlled, applying 8% glycolic acid or 8% L-lactic acid twice a day to the face and outer forearms.

On the face, at least one grade of improvement was reached by 76% on glycolic acid, 71% on lactic acid and 40% on the vehicle.

On the forearm the two parted company. Lactic acid beat the vehicle for roughness; glycolic acid beat it for overall photodamage and sallowness but not for roughness. If you have read that the AHAs are interchangeable, this is the trial that complicates it.

――――――――――――――――――――――――――――

5. Horny layer thickness: urea

A separate study compared urea 2% cream, 10% cream and 10% foam against petrolatum and untreated skin, in healthy volunteers and in people with type I and type II diabetes.

The 10% cream slightly lowered transepidermal water loss and reduced horny layer thickness, with no excessive occlusion from any of the products.

Korea’s safety standards set the ceiling for urea at 10%, so the concentration the study used sits at the regulatory limit rather than below it.

――――――――――――――――――――――――――――

6. The one Korea notifies: salicylic acid 0.5%

Regulation on the Examination of Functional Cosmetics, Annex 4 — acne-prone skin relief

Ingredient and level: salicylic acid 0.5% · Product type: cleansing products of soap composition · Forms: solution, lotion, cream
Claim: “helps relieve acne-prone skin”
Directions: apply an appropriate amount to the skin and rinse off immediately with water

The directions matter as much as the percentage. What the notification covers is a rinse-off cleanser, not a leave-on treatment.

The safety standards then split the ceiling three ways: 0.5% when used as a preservative, 2% in personal cleansing products, and 3% in rinse-off hair products — the latter two only where it is the active of a functional cosmetic.

One more line sits with it: salicylic acid may not be used in products for infants, or in products labelled for children aged 13 or under, with shampoo excepted.

――――――――――――――――――――――――――――

7. Irritation: PHA (lactobionic acid)

Lactobionic acid at 10% was held under a 24 hour occlusive patch in 10 volunteers, with transepidermal water loss, erythema and stratum corneum hydration read afterwards. No irritation appeared.

The hydration arm of the same paper ran in 16 volunteers over 3 hours and 21 days, and there the authors are direct: a statistically significant difference between the placebo and the sample with lactobionic acid was not observed.

Read together, PHA is an ingredient with mildness shown rather than strength shown. That is worth separating when a product page offers it as a gentler route to the same result.

――――――――――――――――――――――――――――

The five side by side

Ingredient	People	Duration	Design	Measured	Result
Glycolic acid 8%	74, shared	22 weeks	Randomised, double blind, vehicle controlled	Facial grade, forearm roughness	Face 76% vs vehicle 40%; no difference in roughness
Lactic acid 8%	74, shared	22 weeks	Same trial	Facial grade, forearm roughness	Face 71%; roughness also beat vehicle
Urea 10%	not in abstract	not in abstract	Compared with petrolatum and untreated skin	Water loss, horny layer thickness	Horny layer thickness fell
Lactobionic acid 10%	10 and 16	24 h and 21 days	Patch test and hydration arm	Irritation, hydration	No irritation; hydration no different from placebo
Salicylic acid 0.5%	—	—	Notified active	—	Acne-prone skin relief

The endpoints are not interchangeable. Roughness, horny layer thickness and irritation are three different measurements, and a product page that says “clinically proven” rarely says which one it means.

And where each study is weak

Study	Where it is weak, stated plainly
Stiller 1996	The facial endpoint is an investigator grade, and roughness was read on the forearm. One participant withdrew for irritation and six for personal reasons.
Borelli 2011	The abstract gives neither the number of participants nor the overall duration. Healthy volunteers and diabetic patients were studied together.
Tasic-Kostov 2011	Hydration showed no significant difference from placebo. The irritation arm is a single 24 hour patch in 10 people.
Salicylic acid	No human trial is attached here — only the level and the directions the notification sets.

――――――――――――――――――――――――――――

8. Korea, the United States and the EU — the same two numbers, different force

No regulator here keeps a claim list for exfoliation. What differs is which instrument fills the gap.

	Korea	United States	European Union
Claim-based ingredient list	Six categories; exfoliation is not one	None — “FDA does not have a list of approved or accepted claims for cosmetics”	Permitted lists cover colourants, preservatives and UV filters (Art. 14)
Test procedure	Ten chapters; exfoliation is not one	Not written down	655/2013 requires “state of the art practices”
AHA and sunlight	Mandatory label text by notification, above 0.5%	Recommended Sunburn Alert, in a guidance	No entry found for this
The same two numbers	Above 10%, or pH below 3.5, adds a line	CIR found 10% or less and pH 3.5 or greater safe	—
Force of the document	Omitting the text is a labelling breach	“do not establish legally enforceable responsibilities”	Common criteria bind; the method is a condition

The striking part is the arithmetic. Korea’s notification adds the physician line above 10% AHA or below pH 3.5. In the United States, the Cosmetic Ingredient Review panel concluded that glycolic and lactic acid are safe where the concentration is 10 percent or less and the final product has a pH of 3.5 or greater. The same two numbers, drawn from the opposite side.

What separates them is the document. Korea’s sits in a notification, so leaving the sentence off the box is a labelling breach. The United States puts its Sunburn Alert wording in a guidance issued in January 2005, and FDA states plainly that its guidances “do not establish legally enforceable responsibilities” and “should be viewed only as recommendations”.

The EU does not run a claim-based ingredient list at all. Article 14 of Regulation (EC) No 1223/2009 keeps positive lists for colourants, preservatives and UV filters, and claims are governed instead by Regulation (EU) No 655/2013: they “shall be supported by adequate and verifiable evidence”, and that evidence “shall take into account state of the art practices”.

For anyone exporting into Korea, the practical reading is short: there is no exfoliation dossier to file, and the compliance work sits on the label instead.

――――――――――――――――――――――――――――

What to check on the box

✔ Ask what was measured — roughness, horny layer thickness or hydration
✔ A sun sensitivity sentence on the box means the product is above 0.5% AHA
✔ For salicylic acid, the ceiling depends on whether the product is rinsed off
✔ Even a small container lists fruit acid (AHA)

――――――――――――――――――――――――――――

One line

Exfoliation is not a reviewed claim in Korea; it is a labelled one. Decide which endpoint matters to you — roughness, horny layer thickness or mildness — and the list of five becomes a list of two.

Next: five actives for skin elasticity — chapter 2 of the same guideline, and unlike exfoliation, a claim with a written test procedure.

――――――――――――――――――――――――――――

What is not in this article, and why

✔ Peel and chemical resurfacing trials — those are procedures, not cosmetics. Only leave-on cosmetic strengths of 8 to 10% were used here
✔ A urea dose-ranging study (Addor et al., Surg Cosmet Dermatol) — the journal page carries no publication year, volume or pages, so it could not be cited properly
✔ A gluconolactone study (PMID 33769633) — Europe PMC returned 429 and PubMed a reCAPTCHA
✔ Enzyme and physical scrubs — no human study was found
✔ Whether Annex III of Regulation 1223/2009 carries entries for glycolic or lactic acid — the annex table could not be opened, so this is recorded as not found rather than as absent

About the author · Dr. J

📌 Related

📌 Acne Skincare Ingredients and Korea’s 0.5% Salicylic Acid Rule — the same salicylic acid entry, read from the acne side

📌 Pores and Sebum: the Test Method the Regulator Wrote Down — the mirror image of today’s piece

📌 Is Glycolic Acid an AHA? AHA vs BHA, Explained — the two acids in today’s trial

📌 Salicylic Acid (BHA): 0.5% vs 2%, Explained — why the ceiling splits

📄 Sources

· Regulation on the Examination of Functional Cosmetics, Annex 4 (Korea Law Information Center, accessed 20 October 2026)

· Regulation on Safety Standards for Cosmetics, Annex 2 (same, accessed 20 October 2026)

· MFDS, Guideline on Test Methods for Substantiating Cosmetic Labelling and Advertising Claims (March 2018, no. 0353-02)

· Notification on Cautions in Use and Labelling of Fragrance Allergens in Cosmetics (cross-checked on the Korea Cosmetic Association ingredient dictionary)

· Enforcement Rule of the Cosmetics Act, Article 19 (Korea Cosmetic Association, statute page)

· Stiller MJ, Bartolone J, Stern R, Smith S, Kollias N, Gillies R, Drake LA. Arch Dermatol 1996;132(6):631-636

· Borelli C, Bielfeldt S, Borelli S, Schaller M, Korting HC. Int J Cosmet Sci 2011;33(1):37-43

· Tasic-Kostov MZ, Reichl S, Lukic MZ, Jaksic IN, Savic SD. Pharmazie 2011;66:862-870

· U.S. FDA, “Alpha Hydroxy Acids”; “Guidance for Industry: Labeling for Cosmetics Containing Alpha Hydroxy Acids” (January 2005); “Cosmetics Labeling Claims”

· Regulation (EC) No 1223/2009, Article 14 · Regulation (EU) No 655/2013, Annex

――――――――――――――――――――――――――――

This article summarises published studies and public regulatory documents; it is not medical advice. The Korean notifications and annexes were opened in the original on the day of writing, and the AHA warning wording was taken from two Korea Cosmetic Association ingredient pages that matched word for word, because the notification annex PDF could not be opened. Of the studies cited, one does not report its participant count and one endpoint showed no difference from placebo; both are stated in the tables. Ingredient names and percentages are given as criteria for choosing a product, not as a recommendation of any particular one.

— Dr. J Ingredient Notes

October 20, 2026

Korea’s Natural Origin Index: The 95% Line, and the Disclaimer That Must Sit Beside the Number (Ingredient Guide 28)

🇰🇷 한국어판

Hello, this is Dr. J. Today we read the natural origin index printed on Korean cosmetic packaging — what that number counts, and why the words “Not a natural cosmetic” so often sit right beside it.

Two things surprise most people. In Korea, “natural cosmetic” is no longer defined in the law. And a product that misses the threshold can still print the index — the rules opened that door on purpose.

――――――――――――――――――――――――――――

The primary sources

📄 Guide on Labelling and Advertising of Natural and Organic Cosmetics (Korea Cosmetic Association, PDF, Korean)

📄 Guideline on Labelling and Advertising of Cosmetics (MFDS, 14 Aug 2025, PDF, Korean)

📄 ISO 16128-2:2017 (abstract)

📄 U.S. FDA, “‘Organic’ Cosmetics”

The Korean documents carry the thresholds and the disclaimer wording; the ISO abstract carries the sentence that limits the standard itself.

――――――――――――――――――――――――――――

Five lines if you are in a hurry

✔ Korea deleted the legal definition of natural and organic cosmetics on 1 August 2025

✔ The thresholds now live in a 2025 industry guide: natural 95%, organic 10% plus natural 95%

✔ A product below the line may still print the index — with the calculation basis shown

✔ Beside that index the front panel carries “Not a natural cosmetic”

✔ ISO 16128 states that it does not address labelling, human safety or environmental safety

――――――――――――――――――――――――――――

The definition left the statute book in August 2025

Until 2025, Korean cosmetics law defined natural cosmetics and organic cosmetics and ran a government certification scheme for them. An amendment that took effect on 1 August 2025 deleted those definitions and the certification provisions together.

What filled the gap is not another statute. It is a guide published in 2025 by the Korea Cosmetic Association — more precisely by the advertising advisory committee of the self-regulatory review body for cosmetic labelling and advertising. The guide describes itself as setting out the minimum requirements for businesses that want to use these words.

※ The repeal itself, and the certification scheme that disappeared with it, were covered in Ingredient Guide 18. This article is about the number that is printed on packs now.

――――――――――――――――――――――――――――

What the guide counts — 95% and 10%

	What the guide says (translated)	How to read it
Natural cosmetic	A product whose natural and natural-origin content reaches 95% or more of the finished product by weight, calculated under the ISO 16128-2 guideline	The unit is the finished product, not an ingredient. Everything in the bottle is weighed
Organic cosmetic	Organic and organic-origin content of 10% or more of the finished product, with natural content — the organic part included — still at 95% or more	Both lines must hold. Organic is a branch of natural here, not an alternative route
What sits on the natural side	Plants and animals, minerals and micro-organisms, and ingredients derived from them	Not plants alone. Mineral and fermentation-derived materials belong here too
The calculation	ISO 16128-1 (definitions for ingredients) · ISO 16128-2 (criteria for ingredients and products)	A two-part international standard, borrowed as the arithmetic

――――――――――――――――――――――――――――

The guide’s scope reaches products that print only an index

This is the clause most summaries skip. The guide does not govern only products that call themselves natural. It also covers products that intend to print nothing but an ISO 16128 index.

Section of the guide	What it sets out
I. Purpose	To give consumers correct information and to set out the minimum requirements for businesses that label and advertise these words
II. Scope	Products labelled or advertised as natural or organic cosmetics, and products that calculate and print a natural or organic (origin) index under the ISO 16128 guideline
III. Criteria	1. Definitions · 2. Labelling and advertising · 3. Substantiation
Substantiation	A product that labels or advertises these words must be able to substantiate that it meets the criteria

――――――――――――――――――――――――――――

Blocked, then opened — the two sentences that explain the pack

Section III-2 of the guide places a prohibition and an exception side by side. Read apart, each one misleads; read together, they explain what you see on the shelf.

	What the guide says (translated)	In practice
The prohibition	Where the finished product does not meet the definition, the product may not directly label or advertise an ingredient as a natural (origin) or organic (origin) ingredient — even if that particular ingredient is one	One natural ingredient does not let the product speak of natural ingredients
The exception	However, where a natural (origin) or organic (origin) index calculated under the ISO 16128 guideline is presented clearly, together with the basis of the calculation, labelling and advertising by index is permitted	The number may be spoken. The calculation basis is the price of admission
The result on the pack	“Natural (origin) index 00% (calculated per ISO 16128)”	Exactly the string you find on Korean packaging

So “natural cosmetic” and “natural origin index” are two different claims. The first belongs to finished products above 95%; the second belongs to any product willing to show its arithmetic.

――――――――――――――――――――――――――――

What the MFDS guideline adds — the four permitted wordings

The Korean Ministry of Food and Drug Safety publishes a Guideline on Labelling and Advertising of Cosmetics (version of 14 August 2025). Its Annex 2, the table of claims that require substantiation, carries a dedicated row: “3. Content concerning ISO natural and organic index labelling and advertising.”

Permitted wording (as printed in the guideline)	Substantiation required
Natural index 00% (ISO 16128 calculation applied)	Substantiated by evidence relating to that finished product “In this case it is also necessary to indicate that the figure is calculated under the ISO 16128 guideline, together with wording that prevents consumer misunderstanding.”
Natural origin index 00% (ISO 16128 calculation applied)	same row
Organic index 00% (ISO 16128 calculation applied)	same row
Organic origin index 00% (ISO 16128 calculation applied)	same row

※ A footnote in the same annex defines the standard plainly: ISO 16128 is a method (guideline) for obtaining a calculated value for natural and organic content in cosmetics.

――――――――――――――――――――――――――――

Why “Not a natural cosmetic” sits beside the number

The guideline’s Section III, Notes is where this article’s hinge lies. When an ISO natural or organic index is labelled or advertised, the guideline says the following ways of writing it are to be followed to reduce consumer misunderstanding.

Where	What must appear alongside (translated)
① Front panel	“Not a natural cosmetic (organic cosmetic)”
② Elsewhere on the pack	“A simple calculation result under ISO 16128; does not mean the product falls under ‘natural cosmetic (organic cosmetic)’”
③ Media advertising	“Natural index 00% (a figure calculated under the ISO 16128 guideline). However, this figure does not mean the product qualifies as a natural (organic) cosmetic”

That small print on the front of the box is not a brand being modest. It is wording the guideline supplies, and the revision history shows when it arrived: the 30 December 2021 edition, recorded as “ISO natural and organic index labelling and advertising method established.”

――――――――――――――――――――――――――――

A prohibition with its own proviso

Annex 1 of the same guideline lists expressions that may not be used. One line reads:

MFDS guideline, Annex 1 — expressions that may not be used

Expressions relating to “natural (Natural)” or “organic (organic)” on a product that does not conform to the Guide on Labelling and Advertising of Natural and Organic Cosmetics (Korea Cosmetic Association).
Note — businesses wishing to use natural or organic expressions need appropriate supporting documents · “provided that expressions falling under Annex 2, item 3 are excluded.”

The last clause carries the weight. The prohibition arrives with its exception attached. A product outside the thresholds cannot say “natural”, yet the index wording of Annex 2 item 3 is carved out of that ban — the same architecture as the guide’s “blocked, then opened” pair above.

――――――――――――――――――――――――――――

What the number does not count — six items, named by the standard itself

The ISO 16128-2 abstract contains one sentence that keeps this whole subject honest.

ISO 16128-2:2017, abstract

“Neither ISO 16128-1 nor this document addresses product communication (e.g. claims and labelling), human safety, environmental safety, socio-economic considerations (e.g. fair trade), characteristics of packaging materials or regulatory requirements applicable for cosmetics.”

The index counts where ingredients came from. Whether a formula is gentle, whether it suits your skin, whether the packaging is sound — none of that is the standard’s business, and it says so.

Which leaves a gap, and the gap is the interesting part: a standard that refuses to speak about labelling is being used for labelling. Each market has filled that silence differently.

――――――――――――――――――――――――――――

One standard, three jobs — Korea, the EU and the United States

	Korea	European Union	United States
Legal definition of natural / organic	Deleted on 1 August 2025. The thresholds now sit in a 2025 industry guide	No definition in the Cosmetics Regulation	FDA does not define it — “The term ‘organic’ is not defined in either of these laws or the regulations that FDA enforces under their authority.”
Where the criteria live	Korea Cosmetic Association guide — natural 95% · organic 10% plus natural 95%	Regulation (EC) No 655/2013 and its six common criteria: legal compliance, truthfulness, evidential support, honesty, fairness, informed decision-making	For organic, the USDA National Organic Program — “The USDA requirements for the use of the term ‘organic’ are separate from the laws and regulations that FDA enforces for cosmetics.”
Status of ISO 16128	Named in the guide and the guideline as the calculation method	Not named in the Regulation	Not named
What must accompany an index	“Not a natural cosmetic (organic cosmetic)” on the front panel	No set wording; the claim must clear the common criteria — “If it is claimed on the product that it contains a specific ingredient, the ingredient shall be deliberately present.”	No set wording
Relationship to safety	Neither the guide nor the guideline is a safety standard	The common criteria govern advertising, not safety	FDA — “An ingredient’s source does not determine its safety.”

※ The six common criteria are the headings of the Annex to Regulation (EC) No 655/2013, applicable since 11 July 2013. Note what all three jurisdictions share: none of them treats a high natural index as a statement about safety.

――――――――――――――――――――――――――――

Four things to check on the box

✔ Look for a percentage next to the word “natural”, not the word alone

✔ Look for “calculated per ISO 16128” — the calculation basis is a condition, not decoration

✔ If you see “Not a natural cosmetic”, that is an index claim — the line is required, not a confession

✔ Judge gentleness from the ingredient list and concentrations, not from the index

――――――――――――――――――――――――――――

In one line

The natural origin index counts where ingredients came from; it does not measure how gentle a product is. Above 95% a product may be called a natural cosmetic; below it, the figure may still be printed with its calculation shown — and the words “Not a natural cosmetic” go with it.

What is not in this article, and why

✔ The calculation formula itself — ISO standards are paid documents. Only the published abstracts were read

✔ The ingredient categories of ISO 16128-1 — same reason; the standard text was not opened

✔ Article and clause numbers of the Enforcement Rule — Korea’s national law portal was unreachable on the day of writing, so no provision numbers are cited

✔ The approval date of the industry guide — the PDF yields only the year (2025)

✔ Certification bodies and marks — none appear in the guide’s text, so none are named here

✔ Claims that natural or organic means gentler — no primary source supports them. The ISO sentence above says the standard does not address human safety, which is the honest version of that question

――――――――――――――――――――――――――――

📌 Earlier articles

📌 How to Read a Korean Cosmetic Ingredient List (Ingredient Guide 14) — checking by name rather than by index

📌 Bakuchiol vs Retinol (Ingredient Guide 3) — the plant-derived side

📌 Collagen: Topical vs Oral (Ingredient Guide 5) — the animal-derived side

📌 Salicylic Acid (BHA): 0.5% vs 2% (Ingredient Guide 8) — who decides a number

About the author · Dr. J

📄 References

· Korea Cosmetic Association, Guide on Labelling and Advertising of Natural and Organic Cosmetics (established 2025)

· Ministry of Food and Drug Safety, Guideline on Labelling and Advertising of Cosmetics, guidance no. 0086-07 (14 August 2025) — Annex 1 · Annex 2 item 3 · Section III Notes

· ISO 16128-1:2016 · ISO 16128-2:2017 (abstracts)

· Cosmetics Act of Korea, Act No. 20767 (in force 1 August 2025) — deletion of the natural and organic definitions and certification provisions

· Regulation (EC) No 655/2013 laying down common criteria for the justification of claims used in relation to cosmetic products

· U.S. FDA, “‘Organic’ Cosmetics” (content current as of 25 February 2022)

――――――――――――――――――――――――――――

This article summarises guidance documents and standard abstracts and is not a substitute for diagnosis or treatment. Passages from the Korean guide and guideline are my translation of the Korean originals; the ISO and FDA sentences are quoted in their published English. Korea’s national law portal was unreachable on the day of writing, so statutory provisions were not re-opened, and ISO standard texts are paid documents, so only the published abstracts were used.

— Dr. J Ingredient Notes

October 13, 2026

Vegan Cosmetics in Korea: The Three Requirements an Industry Guide Set, and the One Line a Korean Label Cannot Print (Ingredient Guide 21)

🇰🇷 한국어판 읽기

Hello, this is Dr. J. A Korean cosmetic box that says vegan almost never says “not tested on animals” beside it. Shoppers read the gap as a gap. It is not one — it is a rule.

Here is the part most summaries skip. No statute anywhere defines the word “vegan” for cosmetics — not in Korea, not in the EU, not in the United States. What differs between the three is what fills that empty space, and what a label is allowed to print next to the word.

――――――――――――――――――――――――――――

Primary sources

📄 Guide to Vegan Labelling and Advertising of Cosmetics, 2025 (Korea Cosmetic Association, PDF, Korean)

📄 Cosmetics Act Article 15-2 (Korean Law Information Center)

📄 Regulation (EU) No 655/2013, common criteria for cosmetic claims (EUR-Lex)

📄 “Cruelty Free”/“Not Tested on Animals” (U.S. FDA)

The Korean guide runs to a few pages and has three chapters: purpose, scope, and criteria. The EU regulation carries the six common criteria in its Annex.

――――――――――――――――――――――――――――

The five lines, if you are in a hurry

✔ “Vegan” is not defined in Korean law — a 2025 industry guide sets the requirements

✔ A product must meet all three: no animal ingredient, none animal-tested, and no testing of the finished product

✔ Ingredients processed with micro-organisms are allowed — fermentation does not break a vegan claim

✔ Share a production line with a non-vegan batch and you wash it thoroughly first

✔ A Korean label cannot print “not tested on animals” — the sale is already banned, so the claim would mislead

――――――――――――――――――――――――――――

The word is not in the statute — a 2025 industry guide fills the gap

Search the Korean Cosmetics Act for “vegan” and nothing comes back. The text that does define it was issued in 2025 by the advertising advisory committee of the self-regulatory review body run by the Korea Cosmetic Association. The guide says of itself that it exists to give advertisers “the minimum requirements for making a vegan claim.”

Its scope is a single line: “This guide applies to cosmetics that are labelled or advertised as vegan.” A product that does not make the claim is outside it.

	What the guide requires	What it means on a shelf
1	A cosmetic that contains no ingredient of animal origin	The ingredient side. Honey, beeswax and lanolin are the familiar examples
2	A cosmetic that contains no ingredient tested on animals under the Cosmetics Act	The ingredient’s history, which never appears on the ingredient list
3	A cosmetic not itself tested on animals under the Cosmetics Act	The finished product, not only its parts
All	“A cosmetic that satisfies all of the requirements below”	Miss one and the word may not be printed

※ The clause numbers inside the guide do not survive text extraction from the PDF, so this article names the chapters rather than the subparagraphs.

――――――――――――――――――――――――――――

Ingredients, manufacturing, substantiation — the three chapters that follow

After the definition the guide sets out how to keep to it. The second line below is the one that surprises formulators.

Chapter	The guide’s wording	Reading
Ingredient standard	“Ingredients of animal origin and ingredients that have been tested on animals shall not be used. An ingredient that has been processed with micro-organisms not derived from animals (bacteria, yeasts, fungi and the like) may be used.”	A fermented ingredient does not break a vegan claim. Using a micro-organism is not the same as using an animal
Manufacturing and control	“Ingredients of animal origin shall be kept from entering a vegan product during manufacture. For example, where a vegan and a non-vegan product are made on the same equipment, machines, equipment, tools and contact surfaces shall be washed thoroughly before the vegan product is made.”	A dedicated line is not required. Cleaning is accepted in its place
Substantiation	“A cosmetic labelled or advertised as vegan shall be able to substantiate that it meets the criteria.”	Making the claim means holding the file. It runs on the same track as Article 14 of the Cosmetics Act

The guide cites, as its legal footing, Articles 10, 13 and 14 of the Cosmetics Act with Articles 19, 22 and 23 of the Enforcement Rule and its Annex 5 — and Article 15-2. The guide itself carries no penalty; those provisions do.

――――――――――――――――――――――――――――

What may be said, and what may not

Two of the four rows below are the ones a brand entering Korea tends to get wrong.

Situation	What the guide sets	On the box
The finished product meets the criteria	It may be labelled and advertised as vegan	The word describes one finished product
Only some ingredients meet them	“A product may not be labelled or advertised as a vegan cosmetic on the basis of particular ingredients alone, rather than the finished product, even where those ingredients meet the criteria.”	“Made with vegan ingredients” and “vegan cosmetic” are different claims
“Not tested on animals”	“Although the distribution and sale of cosmetics tested on animals … is prohibited, wording such as ‘not tested on animals’, which could lead consumers to believe that animal-tested cosmetics are also on the market, may not be used.”	Its absence is not an omission — the phrase is not permitted
Certification marks	The guide carries no rule on certifying bodies or marks	A mark tells you whose standard was met, which you check separately

――――――――――――――――――――――――――――

The statute behind the banned line — Cosmetics Act Article 15-2

The reason the phrase is off-limits sits in the statute. Korea already bans the sale, so printing the claim would suggest that products without it are the tested ones.

Paragraph	The provision	Reading
(1), main text	“A cosmetics responsible distributor or a customised cosmetics seller shall not distribute or sell a cosmetic on which animal testing under Article 2(1) of the Laboratory Animal Act has been conducted, or a cosmetic manufactured (including contract manufacture) or imported using an ingredient on which animal testing has been conducted.”	What is blocked is selling. “Including contract manufacture” sits in brackets, and imports are in the same sentence
(1), proviso	“Provided, that this shall not apply in any of the following cases” (amended 13 Mar 2018, 17 Aug 2021)	Six exceptions follow (next table)
(2)	“The Minister of Food and Drug Safety shall endeavour to develop alternative test methods and shall take the measures necessary for cosmetics responsible distributors and others to use them.”	The duty that pairs with the ban, and the provision that narrows exception 2
Inserted	[This Article Newly Inserted on 3 February 2016]	Reported as in force from 4 February 2017; we could not open the addendum itself

The six exceptions, as the statute words them:

No.	The provision	In short
1	Where necessary to designate standards of use for ingredients requiring particular restriction — preservatives, colorants, ultraviolet filters and the like under Article 8(2) — or to carry out a risk assessment under Article 8(3) of ingredients raising public-health concerns	Safety work. Three ingredient classes are named
2	Where animal testing is necessary because no alternative test method exists (an alternative test method being a method that uses no animals, or that reduces the number of animals used or lessens their pain, recognised as such by the Minister of Food and Drug Safety)	Nothing else will do. The definition sits inside the provision
3	Where animal testing is required by the law of the country of export	The receiving market demands it
4	Where animal testing is required for product development by the law of the country of import	The sourcing market demands it
5	Where an ingredient developed through animal testing under another statute is used in manufacturing a cosmetic	A pharmaceutical-track ingredient, for instance
6	Other cases, designated by the Minister of Food and Drug Safety, where an alternative test is impracticable	Room left for a ministry notice

The penalty is an administrative fine of up to KRW 1 million under Article 40(1)7. Translations of the Korean provisions here are ours.

――――――――――――――――――――――――――――

Korea, the EU and the United States — one word, three situations

Set the three side by side and the pattern is clear. None of them defines the word. What differs is what stands in for a definition, and whether the adjacent claim is allowed.

	Korea	European Union	United States
Is “vegan” defined in law?	No. A 2025 industry guide sets three requirements	No. Regulation (EU) No 655/2013 sets six common criteria for every claim	No. FDA: “there are no legal definitions for these terms”
What fills the gap	A self-regulatory guide, plus a substantiation file	Criterion 6, informed decision-making, and a technical document listing “‘free from animal-derived ingredients’, e.g. in products intended for vegans” among acceptable claims	Nothing. The claim is left to the company
“Not tested on animals”	May not be printed — Article 15-2 already bans the sale	Permitted if the six criteria are met; Regulation (EC) No 1223/2009 Article 18 bans the testing itself	Unrestricted. FDA notes firms “may rely on raw material suppliers or contract laboratories to perform any animal testing necessary”
Who holds the evidence	The advertiser — “shall be able to substantiate”	The responsible person, in the product information file (Article 2 of 655/2013)	The company, by whatever means it chooses

For a brand, the practical reading is that Korea regulates the words around the claim rather than the claim itself. A pack artwork that pairs “vegan” with “not tested on animals” passes in the EU and the United States and fails in Korea — and the Korean reason is not that the statement is false, but that it is already true of everything on the shelf.

The EU technical document is worth one caution. It carries the line “IT IS NOT A EUROPEAN COMMISSION DOCUMENT” and is dated as a version of 3 July 2017, agreed by a sub-working group. It is guidance on how to apply the six criteria, not a statute.

――――――――――――――――――――――――――――

Four things to check on a Korean box

✔ Next to “vegan”, look for which standard or certifier is named

✔ “Made with vegan ingredients” is not the same claim as “vegan cosmetic”

✔ A missing “not tested on animals” is not a missing feature — the phrase is not allowed

✔ Whether a product is gentle is a question of ingredients and concentrations, read separately

――――――――――――――――――――――――――――

In one line

No statute defines “vegan” for cosmetics, and in Korea a 2025 industry guide sets three requirements that must all be met. No ingredient of animal origin, none tested on animals, and no testing of the finished product — with fermented ingredients expressly allowed, and one neighbouring phrase expressly forbidden.

What we left out, and why

✔ The guide’s subparagraph numbers — they do not survive text extraction from the PDF, so we name the chapters instead

✔ Any vegan certifier or mark — the guide names none, and we verified none as a primary source, so we name none either

✔ A list of animal-derived ingredients — the guide carries no such table; honey, beeswax and lanolin appear here as commonly cited examples

✔ The commencement date of Article 15-2 — the law portal does not return addenda, so 4 February 2017 rests on press reports

✔ Any vegan entry in the MFDS labelling guidance — the 14 August 2025 guidance lists natural and organic wording among restricted expressions, but nothing on vegan or animal testing

✔ Claims that vegan products are gentler — no primary source, so not written

――――――――――――――――――――――――――――

📌 Earlier guides

📌 How to Read a Korean Cosmetic Ingredient List (Guide 14) — where the ingredient names are checked

📌 Collagen: Topical vs Oral (Guide 5) — an ingredient that comes from an animal

📌 Bakuchiol vs Retinol — the 44-person trial (Guide 3) — an ingredient that comes from a plant

📌 Sunscreen SPF and PA: How to Read the Numbers (Guide 11) — one of the three classes named in exception 1

About the author · Dr. J

📄 Sources

· Korea Cosmetic Association, Guide to Vegan (VEGAN) Labelling and Advertising of Cosmetics (issued 2025), by the advertising advisory committee of the self-regulatory review body for cosmetic labelling and advertising

· Cosmetics Act of Korea, Article 15-2(1) and (2), Article 40(1)7, Article 8(2) and (3), Articles 10, 13 and 14

· Laboratory Animal Act, Article 2(1)

· Ministry of Food and Drug Safety, Guidance on the Management of Cosmetic Labelling and Advertising (14 August 2025)

· Regulation (EU) No 655/2013, common criteria, Annex points 1–6

· European sub-working group on claims, Technical document on cosmetic claims, Annex III (version of 3 July 2017)

· Regulation (EC) No 1223/2009, Article 18

· U.S. FDA, “Cruelty Free”/“Not Tested on Animals” (content current as of 25 February 2022)

――――――――――――――――――――――――――――

This article summarises an industry guide and statutes for general information and is not a substitute for diagnosis or treatment. On the day it was written the Korean law portal was unreachable, so Article 15-2 was taken verbatim from a statute mirror and the commencement date from press reports. Translations of the Korean provisions and of the guide are ours; the subparagraph numbers inside the guide do not survive text extraction from the PDF, so chapters are named instead.

— Dr. J Ingredient Notes

October 12, 2026

Functional Cosmetics in Korea: The 11 Categories, the Notified Ingredients, and What the Label Certifies (Ingredient Guide 16)

Cover image for a guide to functional cosmetics in Korea, explaining the eleven categories set by the Cosmetics Act and its Enforcement Rule and what the words on the package certify. Ingredient Guide 16 from Dr. J Ingredient Notes.

🇰🇷 한국어판: 기능성화장품 뜻과 종류 11가지 – 무엇을 보증하나

Hello, this is Dr. J. If you have ever turned over a Korean cream and seen the words 기능성화장품 — functional cosmetics — this post is about what that line actually certifies.

It is not a marketing word. It is a legal category, with a fixed list of eleven things a product is allowed to be, a fixed set of ingredients and amounts, and a fixed sentence the package has to carry. Once you know where the line comes from, the shelf reads differently.

There is one more reason this is worth reading if you are outside Korea. The EU forbids exactly what Korea requires. Under EU rules, a cosmetic label may not claim that an authority approved it. In Korea, that approval is the label. Same bottle, two opposite drawers — the table further down sets them side by side.

――――――――――――――――――――――――――――

📄 The original documents

📄 Regulation on the Review of Functional Cosmetics, Annex 4 (Korean)

📄 Enforcement Rule of the Cosmetics Act, Annex 4 — labelling standards (Korean)

📄 21 CFR 333.310 — US acne treatment active ingredients

📄 Commission Regulation (EU) No 655/2013 — common criteria for cosmetic claims

📄 Regulation (EC) No 1223/2009 on cosmetic products

🏠 FDA — Is It a Cosmetic, a Drug, or Both?

The two Korean links download the annex tables themselves. They give the ingredient, the amount, the permitted form, the claim wording and the directions for use, one line each.

――――――――――――――――――――――――――――

The short version

✔ The Act names five groups; the Enforcement Rule splits them into eleven categories

✔ Build a product to a notified ingredient at a notified amount and you file a report instead of a full review

✔ Annex 4 lists 9 whitening ingredients, 4 for wrinkles, 29 UV filters and one for acne

✔ The package carries the words and the reviewed claim, in type at least as large

✔ Hair loss, acne, barrier and stretch marks also carry “not a drug”

――――――――――――――――――――――――――――

Five groups in the Act, eleven categories in the Rule

Article 2(2) of the Cosmetics Act defines functional cosmetics as products falling under one of five items, “as prescribed by Ordinance of the Prime Minister”. That ordinance is the Enforcement Rule, and its Article 2 is where the list becomes eleven.

Cosmetics Act, Article 2(2)	Enforcement Rule, Article 2 — the eleven categories
a. Helps whiten the skin	1. Prevents melanin from settling, so freckles and dark spots are held back 2. Lightens melanin that has already settled
b. Helps improve wrinkles	3. Firms the skin so wrinkles are eased or improved
c. Helps tan the skin evenly, or protects it from UV	4. Shields from sunlight for an even tan 5. Blocks or scatters UV to protect the skin
d. Helps change, remove or nourish hair colour	6. Changes hair colour, or bleaches and strips it 7. Removes body hair (physical removal excluded)
e. Helps prevent or improve dryness, cracking, falling out or keratinization caused by weakened skin or hair function	8. Eases hair loss symptoms (coating and other physical methods excluded) 9. Eases acne-prone skin (cleansing products) 10. Restores barrier function so itching improves 11. Improves the red lines of stretch marks

Read the table left to right and one thing stands out. Four categories came out of a single item — item e. Hair loss, acne, barrier repair and stretch marks all sit under a sentence about dryness, cracking, falling out and keratinization. Those same four are the ones that must carry the “not a drug” sentence, which we will get to.

Anything outside these eleven cannot be sold as a functional cosmetic. A Korean pack may still shout about pores or firmness in large type — that is ordinary advertising copy, not this category.

――――――――――――――――――――――――――――

A review, or a report

Article 4(1) of the Act: “A cosmetics manufacturer, a cosmetics responsible distributor, or a university or research institute prescribed by Ordinance of the Prime Minister that intends to sell a product recognised as a functional cosmetic shall, for each item, either undergo a review by the Minister of Food and Drug Safety as to safety and efficacy, or submit a report to the Minister.”

The review route asks for a good deal. Article 4 of the Regulation on the Review of Functional Cosmetics sets out what has to be filed:

✔ Origin and development history

✔ Safety — single-dose toxicity, skin irritation, sensitisation, photo-toxicity, human patch testing

✔ Efficacy or function — efficacy testing, human application testing, hair-dye efficacy testing

✔ Supporting data for the SPF, water resistance and PA figures

✔ Specifications and test methods, with samples

The report route replaces all of that in three situations: the ingredient, amount, claim, directions, specifications and test methods are already notified; the product is identical in every one of those respects to something already reviewed; or it mixes a reviewed product with a notified one. Article 6 of the same Regulation is the hinge — meet Annex 4 and the data requirement falls away.

――――――――――――――――――――――――――――

Annex 4 — read its title slowly

The table is called “Types of functional cosmetics for which data submission is omitted”. So the amounts in it are not ceilings, and they are not “the dose that works best”. They are the amounts at which you no longer have to prove anything new.

Group in Annex 4	Ingredients	Forms and product type	Claim wording
Even tanning · UV protection	29 filters with maximum levels benzophenone-3 5% · zinc oxide 25% · titanium dioxide 25% and others	Lotion, cream and oil; basic and colour make-up product types	Carried by the SPF and PA figures
Skin whitening	9 ingredients	Lotion, liquid, cream and sheet mask	“Helps whiten the skin”
Wrinkle improvement	4 ingredients	Lotion, liquid, cream and sheet mask	“Helps improve skin wrinkles”
Hair colour change	Dyes, bleaches and strippers	Set out by form in the table	Dyeing, bleaching, stripping
Hair removal	As listed in the table	Liquid, cream, lotion, aerosol	“Depilation (removal of body hair)”
Acne-prone skin relief	Salicylic acid 0.5% — the only one listed	Cleansing product type (soap-based); liquid, lotion, cream	“Helps ease acne-prone skin” Directions: “rinse off immediately with water”

The acne row is the one people misread. The directions column says “rinse off immediately with water”, and the product type is a cleansing preparation. That is why so many Korean products with an acne functional claim are face washes rather than leave-on serums.

Notice also what is not in Annex 4: hair loss, barrier repair and stretch marks. No ingredient has been notified for those three, so a product claiming them goes down the review route. How those individual reviews are decided in practice is something I could not open a source for, so I will leave it there.

――――――――――――――――――――――――――――

The nine and the four, by name

This is the table to hold against a back label. If one of these names appears and the front says functional cosmetic, the product was most likely built to the notified amount and reported.

Whitening (9)	Amount	Wrinkle improvement (4)	Amount
Broussonetia extract	2%	Retinol	2,500 IU/g
Arbutin	2–5%	Retinyl palmitate	10,000 IU/g
Ethyl ascorbyl ether	1–2%	Adenosine	0.04%
Oil-soluble licorice extract	0.05%	Polyethoxylated retinamide	0.05–0.2%
Ascorbyl glucoside	2%	—	—
Magnesium ascorbyl phosphate	3%	—	—
Niacinamide	2–5%	—	—
Alpha-bisabolol	0.5%	—	—
Ascorbyl tetraisopalmitate	2%	—	—

Both groups also have fixed forms: lotion, liquid, cream and sheet mask. The claim wording is fixed too, and so are the directions (“apply an appropriate amount evenly to the skin”; sheet masks are removed after 10 to 20 minutes).

That form restriction has a practical edge. The same ingredient at the same percentage in an essence or ampoule falls outside the table, and the product has to be reviewed rather than reported.

――――――――――――――――――――――――――――

Korea, the United States and the EU — three different drawers

If you buy Korean skincare from abroad, this is the part worth keeping. The same tube can be a state-reviewed “functional cosmetic” in Seoul, an ordinary cosmetic in Paris, and, depending on what the label says, a drug in the United States.

	Korea	United States	European Union
Is there a category for this?	Yes — “functional cosmetics”, defined in the Cosmetics Act	No. A product that claims to affect the structure or function of the body becomes a drug under the FD&C Act	No. Everything is a cosmetic product under Regulation 1223/2009
Who checks before it is sold?	MFDS review, or a report, product by product (Act Article 4)	No premarket approval for cosmetics. OTC drugs may follow a monograph instead	A responsible person must be designated (Art. 4) and the product notified to the Commission (Art. 13)
Where do the numbers come from?	Annex 4 — whitening 9, wrinkles 4, UV filters 29, acne salicylic acid 0.5%	Monograph ranges — for acne, salicylic acid 0.5 to 2% (21 CFR 333.310)	The Annexes restrict what may be used; they do not set amounts for efficacy
May the label say a public authority approved it?	Yes — the words on the pack are exactly that	There is no approval to point to	Not allowed. “Claims that indicate that the product has been authorised or approved by a competent authority within the Union shall not be allowed” (Reg. 655/2013)
What else must the label carry?	The reviewed claim with directions; four categories add “not a drug for preventing or treating disease”	A Drug Facts panel when the product is an OTC drug	Claims must be legal, truthful, evidenced, honest, fair and clear (the six common criteria)

The last two rows are where the systems genuinely collide. The EU’s common criteria open with this: “Claims that indicate that the product has been authorised or approved by a competent authority within the Union shall not be allowed.” The whole point of the Korean line is the opposite — it tells you the Ministry looked at it.

The US draws its line somewhere else again. FDA’s own wording is that cosmetics are “intended to be rubbed, poured, sprinkled, or sprayed on… for cleansing, beautifying, promoting attractiveness, or altering the appearance”, while a drug is “intended to affect the structure or any function of the body”. Claim the second thing and the product changes category. FDA also notes that “cosmeceutical” has no meaning under the law.

So a Korean whitening cream with adenosine at 0.04% is not “approved” anywhere else — the category it belongs to simply does not exist outside Korea. That is worth knowing before you read too much into the words.

――――――――――――――――――――――――――――

What has to be on the package

✔ The words 기능성화장품, or the symbol the Minister designates — Act Article 10(1)8

✔ The claim and the directions that were reviewed or reported

✔ For hair loss, acne, barrier and stretch marks: “not a drug for the prevention or treatment of disease”

✔ For products labelled for infants (3 and under) or children (4 to 13): the preservative content

The placement is fixed as well. Annex 4 of the Enforcement Rule says the sentence goes directly beneath the words 기능성화장품, in type at least as large. So if you can see the words but not the claim, look again just under them.

――――――――――――――――――――――――――――

How I read the label after that

I treat the line as an answer to one question: which of the eleven, and did it clear the procedure for that one? It is a good answer, and it is a narrow one.

It does not tell you how much the product will do. The notified amount is the amount that lets a company skip new data, not the amount chosen for the strongest result. So the next thing I look at is the amount and the form.

For wrinkles, is adenosine there at 0.04%? For whitening, is arbutin or niacinamide in the 2–5% band? And where does that name sit in the ingredient list? Remember that below 1% the order stops meaning anything (Ingredient Guide 14).

――――――――――――――――――――――――――――

Five things to check when buying

✔ Is what you want one of the eleven — whitening, wrinkles, sun, and so on

✔ Does the claim directly under the words match it

✔ Is a notified ingredient in the list — adenosine 0.04%, arbutin 2–5%

✔ Is the form lotion, liquid, cream or sheet mask (for whitening and wrinkles)

✔ For sun care, are the SPF and PA figures printed alongside

――――――――――――――――――――――――――――

One line

“Functional cosmetic” means the product cleared a review or a report for one of eleven named claims. Read the sentence under the words first; then read the amount and the form.

Next up: expiry dates and period-after-opening — the two dates on a Korean package, and what the open-jar symbol means (Ingredient Guide 17).

――――――――――――――――――――――――――――

📌 Earlier posts

📌 Arbutin vs Alpha-Arbutin: What Korea’s 2–5% Means (Ingredient Guide 15)

📌 Adenosine 0.04%: Who Decided That Number (Ingredient Guide 12)

📌 How to Read a Korean Cosmetic Ingredient List (Ingredient Guide 14)

📌 Sunscreen SPF and PA: How to Read the Numbers (Ingredient Guide 11)

📌 Niacinamide: How Much Is Enough (Ingredient Guide 2)

About the author · Dr. J

📄 Sources

· Cosmetics Act (Korea), Articles 2(2), 4 and 10(1)8

· Enforcement Rule of the Cosmetics Act, Articles 2, 9 and 10, and Annex 4 (labelling standards)

· Regulation on the Review of Functional Cosmetics (MFDS Notice No. 2023-61), Articles 4 and 6, Annex 4

· 21 CFR 333.310 — Acne treatment active ingredients (eCFR)

· Regulation (EC) No 1223/2009, Articles 4, 13 and 20

· Commission Regulation (EU) No 655/2013, Annex — common criteria

· US FDA, “Is It a Cosmetic, a Drug, or Both? (Or Is It Soap?)”

――――――――――――――――――――――――――――

This post summarises statutes and regulatory notices and is not a substitute for diagnosis or treatment. If something about your skin worries you, please talk to a professional.

— Dr. J Ingredient Notes

September 30, 2026