"Matrixyl" is the most-cited peptide name in skincare and one of the least precise. It is a trademark of the supplier Sederma that has been attached, over twenty years, to three different peptide products, and the evidence for one is routinely quoted for the others. This page separates them, reads the one randomized trial that exists, reports what a PubMed search returns for the other two, and then explains why the number on a "Matrixyl 10%" bottle has nothing to do with the number in the trial.
Three products under one name
| Trade name | INCI name(s) | What it is | Human evidence isolating it |
|---|---|---|---|
| Matrixyl | Palmitoyl pentapeptide-4 (pal-KTTKS; older name palmitoyl pentapeptide-3) | The five-amino-acid sequence Lys-Thr-Thr-Lys-Ser from collagen I, with a palmitic-acid tail to help it into skin | One randomized placebo-controlled trial (Robinson 2005) |
| Matrixyl 3000 | Palmitoyl tripeptide-1 (pal-GHK) plus palmitoyl tetrapeptide-7 (pal-GQPR) | Two palmitoylated peptides sold as a pair | None found on PubMed on 2026-09-05 as an isolated randomized wrinkle trial; one manufacturer study of a multi-ingredient cream; one supplier methodology paper |
| Matrixyl synthe'6 | Palmitoyl tripeptide-38 | A palmitoylated tripeptide with a modified lysine | Open-label and multi-ingredient studies only |
A product listing "palmitoyl oligopeptide" on an older label is usually palmitoyl tripeptide-1; the INCI name was changed, and the CIR's 2018 review covers both names.
The trial: 93 women, 12 weeks, 3 parts per million
Procter and Gamble's Miami Valley Laboratories ran the one controlled trial of palmitoyl pentapeptide-4 that a reader can check. Ninety-three Caucasian women aged 35 to 55 used, for 12 weeks, "moisturizer control product vs. the same moisturizer product containing 3 ppm pal-KTTKS", one on each side of the face, double-blind and randomized left-right. The peptide "was well tolerated by the skin and provided significant improvement vs. placebo control for reduction in wrinkles/fine lines by both quantitative technical and expert grader image analysis", and subjects reported significant fine-line improvement in self-assessment (Robinson et al., International Journal of Cosmetic Science 2005; the PubMed record carries a 2010 indexing date).
Three ppm is 0.0003%. The trial's design is good: split-face removes most between-person variation, the control is the identical base without the peptide, and the outcome was read by instrument as well as by graders. Its limits are that it is one trial, by the company that sold the product it tested, with no independent replication indexed in the twenty years since. A 2009 analytical paper from the University of Orléans developed a method to assay pal-KTTKS in creams and found that "the influence of the formulation on pal-KTTKS availability is evidenced", which is a polite way of saying that the same labeled amount in two creams need not deliver the same peptide (Chirita et al., Analytica Chimica Acta 2009).
Matrixyl 3000: what a search returns
On 2026-09-05 we searched PubMed for palmitoyl tripeptide-1 together with palmitoyl tetrapeptide-7 in a clinical study. One paper came back: a 2024 12-week study of an eye cream whose "active complex" combined a yeast-rice ferment, N-acetylneuraminic acid and the two peptides, authored by the R&D center of the company that makes the cream, with hydration, elasticity and collagen-density gains reported for the whole product (Yang et al., Skin Research and Technology 2024). It cannot attribute anything to the peptide pair.
Sederma's own 2016 paper is the closest thing to a published test of the blend. It is a methodology study: mass-spectrometry imaging, confocal microscopy and ultrasound were used to map skin proteins by age, and "a blend of palmitoyl oligopeptide and palmitoyl tetrapeptide-7" was applied, with ultrasound showing a thinner, denser sub-epidermal band and confocal microscopy indicating "the peptides improved the ECM structure vs. placebo". The authors describe it as "preliminary" and say further studies will follow (Mondon et al., Journal of Cosmetic Dermatology 2016). A 2021 survey of peptides in sensitive-skin cosmetics, which included palmitoyl tetrapeptide-7, made the general point: "the available data is mostly found in patents and supplier brochures, and not in randomized placebo-controlled studies" (Resende et al., Pharmaceuticals 2021).
The trial figures that retail pages quote for Matrixyl 3000, typically a small volunteer study with a wrinkle-depth percentage, are supplier data. We did not find them in a journal, and we do not repeat them.
Synthe'6: open-label, and never alone
Palmitoyl tripeptide-38 has three human studies we could open, none of them placebo-controlled and none isolating the peptide: a 56-day open-label study of a serum with 15% vitamin C, vitamin E and 5 ppm of the peptide in women over 40, with roughness down 8% to 9% (Lintner et al., Journal of Cosmetic Dermatology 2021, from a formulation consultancy); a 28-day study of a multi-ingredient serum in 32 women by its maker, Isdin (Sanz et al. 2016); and a 12-week open-label study of a multi-ingredient moisturizer in 37 women (Herndon et al., Journal of Drugs in Dermatology 2016). Five ppm, again, is the working concentration in the one study that states it.
Why "Matrixyl 10%" is not 10% peptide
The Ordinary's Matrixyl 10% + HA, $10.90 per 30 mL on 2026-09-05, lists: "Aqua (Water), Glycerin, Butylene Glycol, Palmitoyl Tripeptide-1, Palmitoyl Tetrapeptide-7, Palmitoyl Tripeptide-38, Sodium Hyaluronate ..." and describes the product as containing "two generations of Matrixyl" (theordinary.com). Three things follow. The product contains Matrixyl 3000 and synthe'6, not the palmitoyl pentapeptide-4 of the 2005 trial, so that trial is not evidence for this product. The page does not state a peptide concentration; supplier complexes are dilute solutions of peptide in a carrier, and "10%" of a complex is not 10% peptide. And the working concentrations in the literature are parts per million: 3 ppm in the 2005 trial, 5 ppm in the 2021 study. A label percentage in whole numbers is describing the complex, the carrier, or the marketing, not the peptide.
We are not saying the product is under-dosed or over-dosed; we cannot know, because the number is not published. We are saying the number on the front is not comparable to any number in a paper.
Safety
The Cosmetic Ingredient Review panel's 2018 assessment concluded that tripeptide-1, hexapeptide-12, their metal salts and fatty-acyl derivatives (which include palmitoyl tripeptide-1) and palmitoyl tetrapeptide-7 "are safe in cosmetics in the present practices of use and concentration", on the basis of low use concentrations and negative safety test data (Johnson et al., International Journal of Toxicology 2018). Palmitoyl pentapeptide-4 was taken up separately in a 2023 report (released 2023-08-18 for a September 2023 panel meeting) whose safety tests, as listed in the document, used the peptide at 0.01% and 0.0015%; the draft we opened records the panel's discussion but we could not locate its final conclusion, so we do not state one (CIR Pentapeptides report). The 2005 trial reported the peptide as well tolerated.
What the evidence does not show
It does not show that Matrixyl 3000 or synthe'6 reduce wrinkles on their own; it does not show that any Matrixyl peptide compares with a retinoid; it does not show what concentration is in most products; and it does not show that the 2005 result has been replicated by anyone other than the company that ran it. It does show one well-designed trial in which 3 ppm of palmitoyl pentapeptide-4, in a plain moisturizer, measurably reduced fine lines over 12 weeks.
Sources and dates
Opened 2026-09-05: Robinson et al. 2005 (PubMed 18492182), Chirita et al. 2009 (PubMed 19393372), Mondon et al. 2016 (PubMed 25817264), Yang et al. 2024 (PubMed 38932444), Resende et al. 2021 (PubMed 34451799), Lintner et al. 2021 (PubMed 33103342), Sanz et al. 2016 (PubMed 26424007), Herndon et al. 2016 (PubMed 26151786), the CIR 2018 assessment and 2023 pentapeptides report, and The Ordinary's US product page. PubMed was searched the same day for "palmitoyl tetrapeptide-7 AND palmitoyl tripeptide-1 AND (clinical OR volunteers)" (1 result), "Pal-GHK Pal-GQPR" (5, none a wrinkle trial) and "palmitoyl tripeptide-38" (3). Corrections go to the contact page; the review standard is on the testing notes page.