Palmitoyl Pentapeptide-4

The main human trial, 2005 (PMID: 18492182)

The trial that anchors the ingredient is a 12-week, double-blind, placebo-controlled, split-face, left-right randomized study in 93 Caucasian women aged 35 to 55, published in the International Journal of Cosmetic Science in 2005. The two products were a moisturizer and the same moisturizer containing 3 parts per million pal-KTTKS, that is 0.0003 percent. The paper reports that pal-KTTKS was well tolerated and gave significant improvement against the placebo control for reduction in wrinkles and fine lines, measured both by quantitative technical image analysis and by expert graders, and that subjects also reported significant fine line and wrinkle improvement in self-assessment with directional effects on other parameters. The first author is listed at The Procter and Gamble Company, Miami Valley Laboratories in Cincinnati, and PubMed records no funding statement for the trial. The abstract gives no effect size as a percentage change, so none is quoted here.

Head to head against acetyl hexapeptide-3, 2023 (PMID: 36909866)

A double-blind randomized trial at Kristen Krida Wacana University in Jakarta enrolled 21 Indonesian women aged 26 to 55 and split them into three groups: acetyl hexapeptide-3 cream, palmitoyl pentapeptide-4 cream and placebo, applied twice daily to the periorbital area for eight weeks. Assessment used a Corneometer, Tewameter and Cutometer, digital photography and a Crow's Feet Grading Scale. The authors report that improvements were seen in several subjects in both peptide groups and that palmitoyl pentapeptide-4 appeared to do better than acetyl hexapeptide-3 on the data, the clinical photographs and the self-assessment questionnaire. They call it an initial study and say a larger sample and longer duration are needed. The published abstract reports no p values and no per-group numbers, and no machine-readable full text is available, so this guide reports the direction the authors stated and no statistics. Acetyl hexapeptide-3 is the older INCI name for the compound now called acetyl hexapeptide-8.

Multi-ingredient split-face studies, 2009 (PMID: 19735523)

Two double-blind, randomized, controlled, split-face studies in women aged 30 to 70 with moderate to distinct periorbital wrinkles, 42 subjects in the first and 35 in the second, applied 0.5 g of a product to half the face twice daily for four weeks. The four test products contained niacinamide, the peptides Pal-KT and Pal-KTTKS, and carnosine, and one also contained retinyl propionate. Periorbital roughness was measured by Fast Optical in vivo Topometry. Several of the products significantly improved mean roughness and average maximum roughness against no treatment at four weeks, and a pooled analysis found all three products better than no treatment. Because every product carried several actives, none of this isolates the contribution of pal-KTTKS. The lead author was at Procter and Gamble.

A negative trial in irradiated skin, 2004 (PMID: 15127162)

Twenty women having radiotherapy after breast-conserving surgery were randomized to a cream containing CM Glucan, Hydroxyprolisilan C and Matrixyl or to a dexpanthenol lotion, with skin toxicity scored at 0, 30 and 50 Gy. There was no difference between groups on any toxicity measure, patient satisfaction was high with both, and the only adverse events occurred in the test cream group. The authors concluded that there was no advantage for the test cream and that cost and skin-marking problems argued against recommending it.

The newest human publication is a single case (PMID: 42500768)

A 2026 case report followed one 41-year-old woman for 12 weeks in a split-face design, with a hyaluronic acid, niacinamide and Matrixyl serum on both sides and a red light, galvanic current, vibration and heat device added on one side. Both sides improved on VISIA imaging and the device side improved more. One patient, no control group, and the authors say further studies are needed.

Human skin penetration, measured directly

In excised full-thickness human skin mounted in Franz diffusion cells, microneedle pretreatment raised the fluorescence signal of pal-KTTKS and of two other cosmetic peptides by 2 to 22 fold over passive delivery, which the authors described as the first microneedle permeation data for these peptides (PMID: 25033398). A 2026 study using secondary ion mass spectrometry on tape strips from three healthy women measured the two peptides of Matrixyl 3000, palmitoyl tripeptide-1 and palmitoyl tetrapeptide-7, not pal-KTTKS, and found them only on the outer surfaces of corneocytes with intensity falling steeply with depth (PMID: 42790608).

Animal wound healing (PMID: 35874243)

Forty-two male albino Wistar rats, six per group, were given a dorsal wound and treated daily for 21 days with a Matrixyl cream at 1 mg per mL, a hyaluronic acid patch loaded at 0.1 or 1 mg per mL, a plain patch, a plain cream, a commercial dressing as positive control, or bandaging alone. At day 21 wound area reduction was 90.03 plus or minus 4.42 percent for the 1 mg per mL cream and 94.71 plus or minus 3.04 percent for the 1 mg per mL patch against 69.72 plus or minus 5.17 percent for the untreated group, and histomorphometry put collagen content at 89 plus or minus 3.6 percent for the 1 mg per mL patch against 46.6 plus or minus 3.5 percent for untreated skin. This is a rat study of a wound, not of ageing skin.

Cell and chemistry work

The defining experiment is still the 1993 one: dissecting the carboxyl-terminal propeptide of type I collagen down to residues 212 to 216 and showing that this pentapeptide was the minimum sequence for potent stimulation of collagen and fibronectin production in several mesenchymal cell types, with no change in total protein synthesis (PMID: 8486721). Since then the palmitoylated form has been shown to raise collagen production in human dermal and corneal fibroblasts in a concentration-dependent way (PMID: 23320752), to self-assemble into beta-sheet nanotapes unlike the related palmitoyl peptides (PMID: 23822110), and to be the template for analogues and conjugates that raise collagen in culture (PMID: 27184584; PMID: 28467727; PMID: 24090916; PMID: 19944016; PMID: 25433999). Cell-based screening programmes at Procter and Gamble reported matrix responses to Pal-KTTKS and Pal-KT in human skin equivalent cultures (PMID: 19623777) and listed Pal-KTTKS among the compounds identified by gene expression profiling (PMID: 22670614); the adult-fibroblast result was for a combination of niacinamide, Pal-KTTKS and an olive oil fatty acid derivative rather than for the peptide alone (PMID: 23786618).

What the evidence base does not contain

No trial has measured collagen in human skin biopsies after topical pal-KTTKS. No trial has compared it against tretinoin or retinol. No dose-response trial in people has been published; the 2005 trial tested one concentration, 3 parts per million. The 2011 review's summary, that there is a sound in vitro basis on fibroblasts but relatively few clinical studies and little penetration data, still describes the shape of the literature (PMID: 21535443).

What the trials recorded. The 93-subject, 12-week, placebo-controlled split-face trial reported that pal-KTTKS at 3 parts per million was well tolerated by the skin, and the paper records no adverse effect beyond that statement (PMID: 18492182). The 21-subject 8-week trial in Indonesia set out partly because some anti-aging treatments irritate Asian skin more, and its abstract makes no statement about adverse events (PMID: 36909866). The two split-face studies of multi-ingredient products that contained the peptide make none either (PMID: 19735523).

One human trial with adverse events, in a multi-ingredient cream. A randomized trial in 20 women having radiotherapy after breast-conserving surgery compared a cream whose stated actives were CM Glucan, Hydroxyprolisilan C and Matrixyl against a dexpanthenol lotion. It found no advantage for the test cream, and the adverse events that occurred were all in the test cream group: one suspected allergic reaction and two cases that required the radiotherapy planning simulation to be repeated, with mild itchiness and sporadic efflorescences seen more often (PMID: 15127162). The cream contained three actives, so nothing in that result can be assigned to the peptide alone.

Bioinformatic screening. A 2026 safety framework paper from Procter and Gamble with Karl Lintner as a co-author ran palmitoyl pentapeptide-4 and six other peptides through six bioinformatic tools for toxin and allergen prediction. Palmitoyl pentapeptide-4 showed sequence homology with extracellular matrix proteins of skin, collagen, elastin and fibronectin, and did not raise the toxin or allergen flags that amanitin and conotoxin raised in the same run (PMID: 41953401). That is a computational screen written by people with a commercial interest in the ingredient, not a clinical safety study.

The mechanism-level caution. The compound's job is to increase collagen production, and the one study that looked at the fibrotic side of that found a dose turning point: 0.1 micromolar reduced myofibroblast conversion in culture while 0.5 micromolar did not, and the authors wrote that the balance between wound healing and pro-fibrotic ability should guide dose selection (PMID: 30603464).

No repeat-dose human safety study, no pregnancy data and no data on injected or oral use were found. Everything published is topical or in cell culture.