You are choosing between a copper peptide serum and an injectable vial, or someone told you SNAP-8 works like Botox. Most skin-peptide claims start with cells in a dish. This page shows which compounds were tested on human skin, by which route, and what the trials found.
The dot marks each compound's grade, from A (approved drug or several randomized trials) to D (cells and animals only). Rows above the line have some human data. Rows below it have none.
FDA-approved as the Scenesse implant for a rare disorder where light on the skin causes pain. Not approved for tanning.
FDA-approvedThe copper peptide in serums. Wound-gel trials split, the wrinkle-cream data were never published in a journal, and injection was never tested.
Research chemical WADA prohibitedYour own skin makes it. Tested as an ulcer cream in three placebo trials; the largest missed its main goal.
Research chemical WADA prohibitedA longer version of Argireline sold for expression lines. Its own human data are a supplier test and mixed-ingredient patches.
Research chemicalSold for hair growth. The evidence is one study of human hair follicles kept alive in a dish.
Research chemical WADA prohibitedA three-amino-acid piece of alpha-MSH that calmed skin and gut inflammation in mice. No human study of any kind.
Research chemical WADA prohibitedThe doses used in each published trial are on every peptide page for free members. Create a free account
The same peptide can have a trial as a cream and nothing at all as an injection. Each column is a level of evidence. Tap any entry for the study, who was in it, and the source.
4 of 7 compounds on the skin have a randomized human trial for a skin or hair use.
None
Matrixyl and Argireline are cosmetic ingredients with no peptide profile on this site, so they have no grade or legal status. Argireline is the parent of SNAP-8; its trial is listed on the SNAP-8 page.
The outer layer of skin, the stratum corneum, is a wall of flattened dead cells sealed with fat. Bos and Meinardi pointed out in 2000 that nearly every drug that works through skin, and nearly every common contact allergen, weighs less than 500 daltons. Larger molecules mostly stay on the surface.
Peptides are chains of amino acids at roughly 110 daltons each, so anything longer than four or five links is over that line. The record for KPV shows what that means in practice: in FDA's 2026 review it did not pass the outer layer of human cadaver skin unless the skin was microneedled first.
That is why the trials that worked used whole products applied for weeks, and why sellers push injection. Injection skips the barrier, but none of the cosmetic peptides here has been tested that way in people.
Step through each mechanism on a cross-section of skin. Every step is tagged with where it was shown: in cells, in animals, or in people. Watch where the chain breaks.
The GHK sequence sits inside collagen and SPARC. Enzymes at an injury can cut it loose, so GHK may act as a local repair signal.
No trial of injected GHK-Cu has been published, for skin, hair or anything else. That includes its safety, how long it lasts and whether antibodies form.
AHK-Cu lengthened follicles in a dish. No study has put it, or GHK-Cu, on or into a human scalp and counted hairs.
The only SNAP-8 data are supplier tests and patches mixed with other actives. The 60-person trial was for its parent, Argireline.
Argireline and SNAP-8 are meant to quiet nerve signals to facial muscle. A 2025 review concluded it is still uncertain whether enough crosses skin to do that.
No carcinogenicity studies were done for the Scenesse label. Melanotan-2 users have case reports of new and darkening moles, and melanoma has been reported, though nobody has shown the drug caused it.
None of the studies tested the powders sold online. LL-37 is 37 amino acids long and harder to make cleanly than short peptides.
No peptide has been shown to regrow hair on a human scalp. AHK-Cu lengthened human follicles kept alive in a lab, and copper peptides enlarged follicles in mice, rats and macaques. Nobody has published a scalp trial with hair counts.
The approved hair-loss drugs, minoxidil and finasteride, are not peptides. A dermatologist can tell you whether either fits your type of hair loss.
AHK-Cu profileAfamelanotide (Melanotan-1) darkens skin without sun by switching on the MC1R pigment receptor. The approved implant is only for erythropoietic protoporphyria. Swallowed, it did not reach the blood at all.
Melanotan-2 is a different molecule that was never approved. Its controlled human data are a few small trials from 1996 to 2000. Later case reports in people using unregulated vials describe new and darkening moles, kidney injury and melanoma, without proof that the drug caused the cancer.
| Peptide | Evidence | Routes on record | Legal status | Human studies |
|---|---|---|---|---|
| Melanotan-1 (afamelanotide)Melanocortin | A Evidence | Subcutaneous implant, Subcutaneous injection | FDA-approved | 25 |
| GHK-CuTissue repair | B Evidence | Topical, Subcutaneous injection | Research chemical WADA prohibited | 6 |
| LL-37Immunomodulator | B Evidence | Topical, Intratumoral injection, Subcutaneous injection | Research chemical WADA prohibited | 4 |
| SNAP-8Cosmetic / topical | C Evidence | Topical, Microneedle patches | Research chemical | 3 |
| AHK-CuCosmetic / topical | D Evidence | Topical, Subcutaneous injection | Research chemical WADA prohibited | 0 |
| KPVImmunomodulator | D Evidence | Oral, Subcutaneous injection, Topical | Research chemical WADA prohibited | 0 |
A 210-patient phase 3 trial reads out at the end of 2026. One email when it does, and whenever a skin-peptide trial publishes or the FDA changes a peptide's status, with the source linked.