Korea gives you 15 days to prove a claim
Most brands lock the formula first and write the product page later. In Korea those two steps are one. The MFDS can ask you to substantiate any factual claim you have published, and the clock it starts runs for fifteen days from the day the request arrives.
In Korea you may claim only as much as your evidence covers. Article 14 of the Cosmetics Act lets the MFDS demand substantiation for any factual claim in your labelling or advertising, and the data is due within 15 days of the request. Keep advertising without filing it and the MFDS must order the claim pulled until the data arrives.
You may claim as much as your file covers
Buyers sourcing from Korea read the formula sheet and the unit cost closely. The sentences that will sit on the product page usually get written at the end, by the marketing team, in another country. Korea treats those sentences as part of the product. Article 14 of the Cosmetics Act requires the responsible company to be able to substantiate anything factual it has said in labelling or advertising.
Substantiation is not a dossier you file in advance. The MFDS picks out a claim it considers likely to mislead consumers, names it, and asks for the data. From the day that request lands you have 15 days, extendable only where the ministry accepts that there is good reason.
Fifteen days is not enough to run a study
This is where overseas brands get caught. No panel recruitment, test-institute booking and measurement cycle fits inside two weeks. So in practice the deadline means something stricter than it sounds: the evidence has to exist on the day the claim goes live.
Keep the claim running without filing and the ministry is required to order it suspended until the data is submitted. Ignoring that order carries up to one year in prison or a fine of up to 10 million won, and the two can be imposed together. Trimming a sentence in advance is far cheaper than pulling a live listing.
The evidence differs by wording
Which claims are substantiation targets, and with what, comes from the MFDS notification on substantiation of cosmetic labelling and advertising. The Korea Cosmetic Association's 2025 advertising review handbook lays the pairings out:
| Claim | Evidence required |
|---|---|
| Suitable for acne-prone skin | Human application study |
| Controls sebum production | Human application study |
| Blocks / prevents adhesion of fine dust | Human application study |
| Reduces the look of dark circles or puffiness | Human application study |
| Antibacterial (rinse-off cleansing only) | Human application study |
| Improves hair damage | Human application or in-vitro study |
| Anti-ageing, reduces signs of ageing | Human application or in-vitro study |
| Numeric results such as "30% more hydration" | Human application or in-vitro study |
| "Free of X" | Analytical test report |
The first five rows accept nothing else. A cell assay or a supplier brochure is not an answer to that request, which is worth knowing before the copy is signed off. The hair and ageing wordings have two routes, and that changes what the testing budget has to cover.
Some wordings ride on the functional dossier
If the product has already been through functional-cosmetic review or notification, that dossier does the work. Melasma and freckle wording draws on the whitening file, hair-fall wording on the hair-loss relief file. Broader anti-ageing language can also be covered, provided the claim runs through a recognised functional efficacy such as UV protection or wrinkle improvement.
Mechanism language is the exception. "Increases collagen" or "activates enzymes" has to appear inside the wrinkle-improvement dossier, or be substantiated separately. Most of the trouble here starts with a diagram lifted from an ingredient supplier's deck and pasted onto a product page, where it stops being a supplier's statement and becomes yours.
Add one line to the sample request
The fix is an order change. Instead of writing copy once the formula is locked, send your manufacturer five candidate sentences up front and ask which evidence each one needs. That is also the stage where an existing base with a completed human study becomes visible as an option, which can be the difference between a claim you can afford and one you cannot.
Two wording details travel with this. Korean advertising uses "human application study" rather than "clinical", and the test institute, test subjects and test period are expected to appear alongside the figure. Where a sentence sits on the line, the association runs a pre-review service, and a conditional result can be re-reviewed once within a month at no fee.
See which ingredients are moving right now on today's ingredient signals, and check category price bands and cost structure in the product planner.
In short
- Substantiation is filed on request, not in advance. The deadline is 15 days from the request, and extensions need accepted cause.
- No new study finishes inside that window, so the data has to be in hand when the claim is published.
- Acne-suitability, sebum control and fine-dust claims take a human application study only. Hair damage and ageing wordings also accept in-vitro data.
- Export-only products may follow the importing country's rules under Article 30, but a reverse cross-border listing into Korea pulls the Korean standard back in.
*Sources: Korea Ministry of Government Legislation, Easy Law — "Substantiation of cosmetic labelling and advertising" (Cosmetics Act Art. 14; Enforcement Rule Art. 23; Arts. 30 and 37), accessed 18 October 2026 / MFDS Notification No. 2020-80, Regulation on Substantiation of Cosmetic Labelling and Advertising, 4 September 2020 / Korea Cosmetic Association, Cosmetic Advertising Review Criteria (Commentary), 2025*
Questions people ask
Only with a human application study behind it. The same applies to "controls sebum production", "blocks fine dust" and, for rinse-off cleansing products, "antibacterial". An in-vitro study does not substitute for any of them.
For some wordings the review or notification dossier is the evidence. "Helps reduce the look of melasma and freckles" rests on the whitening dossier, "reduces hair fall" on the hair-loss relief dossier. A mechanism claim such as "increases collagen" is different: it must either sit inside the wrinkle-improvement dossier or be substantiated on its own.
An analytical test report. Only where presence cannot be verified by testing may you fall back on manufacturing records or a raw-material certificate of analysis. Ingredients already banned from cosmetics are outside this list.
Yes, if you name the comparison and the basis, and the point can be verified objectively. The evidence is a human application or in-vitro study. Leaving out what the product was compared against is itself the violation.
Article 30 lets a product made solely for export follow the importing country's rules instead. The moment it is also sold in Korea, including through a reverse cross-border storefront, the Korean standard applies in full.