Peptide Glow Journal

Peptides for Skin: The Whole Category, Counted

We counted every randomized trial behind the peptides on skincare labels: nine for topical peptides, fifty-five for retinoids. The census, and what it means.

Peptide Glow Journal Editorial · Published 2026-09-06

Every serum on the shelf now lists a peptide, and almost every one of them borrows its credibility from the same handful of studies. So we counted. On 2026-09-06 we ran the trial census on PubMed: how many randomized controlled trials exist for topical peptides against wrinkles and photoaging, how many for each named peptide, and how many for the benchmark the category compares itself to. The answers are small numbers, and the numbers are the story.

The census

PubMed, searched 2026-09-06, title and abstract terms, publication type "randomized controlled trial". Anyone can re-run these and get a number; that is the point of publishing them.

Ingredient Randomized controlled trials indexed Human records of any kind
Retinoids (tretinoin, retinol) in photoaging/wrinkles 55
Oral collagen peptides, skin outcomes 38 188
Topical peptides, wrinkles or photoaging (all) 9
Acetyl hexapeptide-8 (Argireline) 2 (one cosmetic, one blepharospasm) 22
Palmitoyl pentapeptide-4 (Matrixyl) 1 (radiotherapy skin care, not wrinkles) 10
Copper tripeptide / GHK-Cu 1 (negative on objective endpoints) 47
Acetyl octapeptide-3 (SNAP-8) 0 0

Two figures deserve their own sentence. The whole topical-peptide category, every peptide, every trial, has nine randomized trials against wrinkles or photoaging. And the thing the category is most often compared to has fifty-five.

What those nine trials actually are

A count is only useful if you read what you counted. The nine:

  • A cyclized hexapeptide-9 against retinol — randomized, double-blinded, active- and vehicle-controlled, 56 days, 0.002% of each. The peptide beat 0.002% retinol on crow's-feet area and roughness and on forehead wrinkle number, area and roughness. Authors: the Global R&D Center of Shanghai Chicmax Cosmetic Co. and Shanghai Kans Biotech, which develop and sell it (Chang et al., J Cosmet Dermatol 2025).
  • A cosmetic regimen against 0.02% tretinoin — 196 women, 8 weeks, with 25-person cohorts continuing to 24 weeks. The cosmetic regimen improved wrinkle appearance significantly more than tretinoin at 8 weeks and comparably at 24, with better tolerance. But it was a regimen of three products containing 5% niacinamide, peptides, antioxidants and 0.3% retinyl propionate, so nothing in it can be attributed to the peptides. Authors: Procter & Gamble (Fu et al., Br J Dermatol 2010).
  • Argireline in Chinese subjects — randomized, placebo-controlled, the one indexed cosmetic wrinkle trial of a single named peptide (Am J Clin Dermatol 2013); covered in detail on our Argireline page.
  • A repair-of-photoaged-dermal-matrix study of a cosmetic anti-ageing product (Br J Dermatol 2008) — a product, not an isolated peptide.
  • Five others that test a matrikine-like micro-protein complex, an AI-discovered natural peptide, a zeaxanthin supplement plus topical serum, a post-laser wound-healing peptide, and a collagen injection combined with a peptide serum. Every one is a multi-component product or a non-cosmetic endpoint.

Read that way, the number of randomized trials that isolate a single named cosmetic peptide and measure wrinkles is roughly two, depending on how generously you count the 2025 hexapeptide-9 study, whose sponsor made the peptide.

The copper trial nobody cites

Copper peptides are the most-discussed ingredient in this category, and there is exactly one randomized trial of one indexed on PubMed. Thirteen patients had circumoral CO2 laser resurfacing and were randomized to a post-treatment regimen with or without GHK-Cu. Blinded evaluators and computer image analysis found no significant difference in the resolution of erythema; objective assessment found no significant improvement in wrinkles or overall skin quality. The only endpoint that separated the groups was the patients' own questionnaire (P = .04) (Miller et al., Arch Facial Plast Surg 2006).

The trial that consumer articles do cite — a 40-woman randomized double-blind study reporting wrinkle volume down 24.1% against 15.0% for vehicle — was funded by the serum's maker and published in a journal PubMed does not index. Both are real studies. One is indexed, negative on its objective endpoints, and almost never mentioned; the other is sponsor-funded, positive, and everywhere. Our copper peptides page has the full table.

The molecules, and whether they get in

Peptides on labels fall into four working groups, and the grouping is mechanistic marketing rather than a regulatory classification:

  • Signal peptides — palmitoyl pentapeptide-4 (Matrixyl), palmitoyl tripeptide-1 — claimed to prompt fibroblasts toward collagen.
  • Neurotransmitter-inhibiting peptides — acetyl hexapeptide-8 (Argireline), acetyl octapeptide-3 (SNAP-8) — claimed to interfere with the SNARE complex, the "topical botulinum" story.
  • Carrier peptides — GHK-Cu — which deliver a metal ion, in this case copper, as well as acting themselves.
  • Enzyme-inhibitor peptides — soy and rice-derived sequences claimed to slow matrix breakdown.

All four run into the same wall. The 500-dalton rule holds that compounds much above 500 Da do not passively cross intact stratum corneum in useful amounts. GHK-Cu is about 403 Da and clears it; acetyl hexapeptide-8 at about 889 Da and palmitoyl pentapeptide-4 at about 578 Da do not. Peptides are also charged and water-loving, which is the opposite of what crosses a lipid barrier. That is why formulators palmitoylate them, and why the one direct measurement we could open — an FDA penetration study of acetyl hexapeptide-8 — found 0.22% of a 10% dose in the stratum corneum and 0.01% reaching the epidermis after 24 hours.

The percentages are not comparable

A "10%" on two peptide bottles can mean concentrations 2,000-fold apart, because suppliers sell some peptides as dilute solutions and the label may count the solution:

Label What it usually means Peptide actually present
"10% Argireline" 10% of the peptide ~10% acetyl hexapeptide-8
"10% SNAP-8" 10% of a 0.05% supplier solution ~0.005% acetyl octapeptide-3
"1% copper peptides" (The Ordinary) 1% copper tripeptide-1 1%
The 2025 trial that beat retinol 0.002% cyclized hexapeptide-9 0.002%

The last row is the one worth keeping. The peptide that outperformed retinol in a randomized trial did it at two-thousandths of a percent. No published dose-response curve exists for any cosmetic peptide, so a larger number on a label is a marketing decision, not a measured advantage.

What this evidence supports, and what it does not

It supports: modest, instrument-measured improvements in wrinkle appearance from some peptides in small short trials; excellent tolerance compared with retinoids; and a genuine mechanistic literature, mostly in cell culture, behind the claims. It does not support: peptides as equivalent to retinoids, ranked comparisons between peptides, any claim that a peptide rebuilds or regenerates skin, or the idea that a higher label percentage buys more effect. Under US law a cosmetic is an article for "cleansing, beautifying, promoting attractiveness, or altering the appearance", and appearance is what these trials measured.

Where to go next

Ingredient by ingredient: Argireline, Matrixyl, SNAP-8, and copper peptides — with the layering question, and the two laboratory papers behind it, in what not to mix with copper peptides. Ranked by evidence rather than by marketing: best peptides for skin. Swallowed rather than applied: collagen peptides. If your interest is hair, the evidence is a different set of peptides entirely — peptides for hair growth and why the same ingredient rarely has evidence for both. The injectable end of the category is on our glow stack page. Our testing notes explain the standard this publication holds itself to.

Sources

  • PubMed census run 2026-09-06 via the NCBI E-utilities esearch endpoint, publication type randomized controlled trial[pt], terms as quoted in the table. PubMed.
  • Chang H, Tao K, Yang Y, et al. Novel Cyclized Hexapeptide-9 Outperforms Retinol Against Skin Aging: A Randomized, Double-Blinded, Active- and Vehicle-Controlled Clinical Trial. J Cosmet Dermatol 2025;24(7):e70290. PubMed 40586182. Authors at Shanghai Chicmax Cosmetic Co. and Shanghai Kans Biotech.
  • Fu JJ, Hillebrand GG, Raleigh P, et al. A randomized, controlled comparative study of the wrinkle reduction benefits of a cosmetic niacinamide/peptide/retinyl propionate product regimen vs. a prescription 0.02% tretinoin product regimen. Br J Dermatol 2010;162(3):647-54. PubMed 20374604. Authors at The Procter & Gamble Company.
  • Miller TR, Wagner JD, Baack BR, Eisbach KJ. Effects of topical copper tripeptide complex on CO2 laser-resurfaced skin. Arch Facial Plast Surg 2006;8(4):252-9. PubMed 16847171.
  • Wang Y, et al. The anti-wrinkle efficacy of argireline, a synthetic hexapeptide, in Chinese subjects: a randomized, placebo-controlled study. Am J Clin Dermatol 2013;14(2):147-53. PubMed 23417317.
  • Watson RE, et al. Repair of photoaged dermal matrix by topical application of a cosmetic 'antiageing' product. Br J Dermatol 2008;158(3):472-7. PubMed 18070204.
  • Bos JD, Meinardi MM. The 500 Dalton rule for the skin penetration of chemical compounds and drugs. Exp Dermatol 2000;9(3):165-9. PubMed 10839713.
  • Gorouhi F, Maibach HI. Role of topical peptides in preventing or treating aged skin. Int J Cosmet Sci 2009;31(5):327-45. PubMed 19570099.
  • 21 U.S.C. § 321(g)(1) and (i), Cornell LII.
  • Penetration and concentration figures as sourced on our Argireline, Matrixyl, SNAP-8 and copper peptides pages, each citing the primary document.

Published by Peptide Glow Journal Editorial. No product is ranked, scored or affiliate-linked on this page.

Frequently asked questions

Do peptides actually work for skin?

Some have measured effects in small trials; the category as a whole has far less evidence than its marketing implies. Counted on 2026-09-06, PubMed holds 9 randomized controlled trials of topical peptides for wrinkles or photoaging, and most of those test a finished product with several actives rather than a peptide on its own. Retinoids have 55. The best-evidenced peptide result is the 2013 placebo-controlled Argireline trial in Chinese subjects; the most recent is a 2025 trial of a cyclized hexapeptide-9 that beat retinol at the same 0.002% concentration, run by the peptide's own developer.

Which peptide has the most evidence for skin?

If you count trials rather than claims: oral collagen peptides, with 38 randomized trials indexed for skin outcomes and two meta-analyses behind them, but that is a supplement you swallow, not a serum. Among topical peptides, acetyl hexapeptide-8 (Argireline) has the only randomized placebo-controlled cosmetic wrinkle trial of a named, single peptide that we could find indexed. Copper peptides have a large mechanistic literature and one indexed randomized trial, which was negative on its objective endpoints.

Are peptides better than retinol?

The evidence bases are not comparable in size: 55 randomized trials for retinoids in photoaging against 9 for topical peptides, and the peptide trials are smaller and more often run by the ingredient's maker. One 2025 randomized, double-blinded trial did report a cyclized hexapeptide-9 outperforming retinol at 0.002% on crow's feet and forehead wrinkles, and one 2010 trial found a cosmetic regimen containing peptides matched 0.02% tretinoin at 8 weeks with better tolerance. Both are real results; neither makes the category's evidence equal to the retinoid literature. Peptides are generally better tolerated, which is a real advantage for people who cannot use retinoids.

Do peptides even get into skin?

That is the category's unresolved problem. The 500-dalton rule holds that compounds above roughly 500 Da do not passively cross intact stratum corneum well; GHK-Cu is about 403 Da, acetyl hexapeptide-8 about 889 Da, palmitoyl pentapeptide-4 about 578 Da. The one direct measurement we could open, an FDA penetration study of acetyl hexapeptide-8, found 0.22% of a 10% dose in the stratum corneum and 0.01% in the epidermis after 24 hours. Palmitoylation exists to solve exactly this, and formulators use carriers and delivery systems for the same reason.

What do the percentages on peptide labels mean?

Often not what they look like. Suppliers sell peptides as dilute solutions, and a label percentage may count the supplier solution rather than the peptide. A product labelled 10% SNAP-8 contains about 0.005% acetyl octapeptide-3, because the raw material is a 0.05% solution. The Ordinary's copper serum lists copper tripeptide-1 at 1%. The 2025 hexapeptide-9 trial that beat retinol used 0.002%. There is no established dose-response curve for any cosmetic peptide, so a bigger number on the front of a bottle is not evidence of a bigger effect.

Which peptides should you avoid?

This page does not tell anyone what to use or avoid. What the record supports saying is which ingredients have no published human evidence at all: acetyl octapeptide-3 has zero human records on PubMed, and several peptides marketed for skin have only their supplier's own unpublished testing. Injectable research compounds sold for appearance are a separate matter with separate risks, covered on our glow-stack page.

How long do peptides take to work?

The trials that exist measured at 28 days (supplier tests), 8 weeks (the Argireline trial, the copper serum trial, the 2010 regimen comparison at its first endpoint) and 56 days (the 2025 hexapeptide-9 trial). Almost nothing runs past 12 weeks, and the longest arm we found in this literature is the 24-week continuation of the 2010 tretinoin comparison, in cohorts of 25.