Three unrelated things are sold as skin peptides: ones you swallow (hydrolysed collagen, meaning collagen broken into fragments small enough to absorb), ones you rub on (cosmetic ingredients such as Matrixyl and the copper peptide GHK-Cu), and ones you inject (BPC-157, epithalon). The swallowed ones have by far the most human evidence, with 26 randomised trials and 1,721 people pooled for skin water content and stretchiness, which makes the cheapest category in this market the only one with a real trial base and the most expensive category the one with none. Even there the effects are modest instrument readings rather than a visible change in the mirror, and the research is heavily funded by the industry that sells it. Rubbed-on peptides have a handful of small trials, several run by the company selling the ingredient, and the copper peptide's only independent randomised facial trial found no objective benefit. Injected peptides for skin have no randomised human skin trial at all and are not FDA-approved for human use.
Three unrelated things get sold as peptides for skin, and the category with the most human trial evidence behind it is the cheapest and least interesting of the three.
Sort the market by how the compound gets into you and the confusion clears. Peptides you swallow are hydrolysed collagen, meaning collagen broken into fragments small enough for your gut to absorb, sold as a powder or a capsule. Peptides you rub on are cosmetic ingredients in a serum or a moisturiser: palmitoyl pentapeptide-4, sold as Matrixyl, acetyl hexapeptide-3, and the copper peptide GHK-Cu. Peptides you inject are compounds such as BPC-157 and epithalon, bought from research-chemical suppliers as a powder in a vial. The three share a word and very little else.
Which skin peptides have human trials behind them?
The table below is the whole argument in one place: what each category is, what was measured, in how many people, what it costs relative to the others, how strong the design was, where it sits legally in the US, and which page carries the compound detail.
| Category | Examples | What was measured | People pooled | Relative cost | Design | Status in the US | Detail |
|---|---|---|---|---|---|---|---|
| Swallowed | Hydrolysed collagen peptides | Skin water content and stretchiness by instrument, plus wrinkle scores | 1,721 across 26 trials 2; 1,125 across 19 more 3 | Cheapest of the three categories per month | Randomised and placebo-controlled, pooled in two independent reviews, heavily industry funded | Dietary supplement, not a medicine | Oral bioavailability |
| Rubbed on | Palmitoyl pentapeptide-4 (Matrixyl), acetyl hexapeptide-3, GHK-Cu | Fine lines and wrinkles by image analysis and trained graders | 93 in the largest single trial 4; 21 in the only head-to-head 5; 13 in the only independent GHK-Cu facial trial 6 | Mid-priced, and the peptide concentration is often not stated on the label | Randomised and blinded, but small, short, and several run by the company selling the ingredient | Cosmetic ingredients; none is an FDA-approved drug for any skin use. GHK-Cu sold in vials for injection is not a cosmetic, it is an unapproved drug sold for research use | GHK-Cu |
| Injected | BPC-157, epithalon, TB-500 | Nothing, on skin | 0 | The most expensive, and the only category with nothing in the People pooled column | No randomised human skin trial has been published 9 | Unapproved investigational compounds, not FDA-approved for human use; removed from FDA 503A Category 2 in April 2026, which is not approval and does not authorise compounding | BPC-157 |
Read the People pooled column against the Relative cost column: the evidence ranking runs close to the reverse of the price ranking. The cards below cover the compounds named on this page plus TB-500, which is sold into the same market and, like the rest of the injected category, has no human skin trial behind it.
Does swallowed collagen work for skin?
Two independent pooled analyses say yes, with caveats that belong in the same breath. The 2023 review in Nutrients pooled 26 randomised controlled trials and 1,721 patients and found that hydrolysed collagen improved skin water content and stretchiness against placebo, both comfortably past the usual statistical threshold (Z = 4.94 and Z = 4.49, both p < 0.00001, where a p-value is the chance of seeing a result this big if the product did nothing, so smaller is stronger) 2. A 2021 review in the International Journal of Dermatology pooled 19 randomised, double-blind, controlled trials and 1,125 participants aged 20 to 70, reporting favourable results for water content, stretchiness and wrinkles after roughly 90 days of daily ingestion 3.
What was measured is mostly an instrument reading: a probe that records how much water the skin surface holds, and a device that measures how quickly skin snaps back when gentle suction is applied. Those are real measurements, and they are a different thing from looking different in a mirror.
A different set of 19 randomised trials, 1,341 participants, pooled in a 2026 systematic review that included rubbed-on peptides alongside swallowed ones 1, landed in the same place. Peptides improved skin water content and brightness, the pooled effect on wrinkles was modest (mean difference 0.27, p = 0.04), and a subgroup analysis found that the benefit was driven largely by the swallowed ones (mean difference 1.5, p = 0.01). Effects on stretchiness and skin density were inconsistent. The review also reports that peptides were well tolerated with minimal adverse events across the trials it pooled, and unlike most of this literature it is not a manufacturer's study, though its no-commercial-relationships line is that journal's standard declaration rather than an unusual one.
Whether a swallowed peptide survives the gut intact is a separate question with its own answer, covered on our page on oral peptide absorption. The swallowed category comes out of all this with small effects measured by instruments, a funding problem running through its literature, and the distinction of being the only one of the three categories where somebody ran the trials.
Do peptide serums work?
The largest single trial of a cosmetic skin peptide ran 93 women aged 35 to 55 for 12 weeks, split-face and double-blind, meaning each woman got the peptide moisturiser on one side of her face and the plain one on the other, so she was her own comparison. It tested a moisturiser containing 3 parts per million of palmitoyl pentapeptide-4, the ingredient sold as Matrixyl, against the identical moisturiser without it, and found significant improvement in fine lines and wrinkles by image analysis and by expert grading 4. The published paper carries two details the quoted version tends to leave out. The comparison was the base moisturiser rather than retinol, and the authors' affiliation is The Procter and Gamble Company, Miami Valley Laboratories.
The only trial we could find putting two cosmetic peptide creams against each other, acetyl hexapeptide-3 against palmitoyl pentapeptide-4 for crow's feet, enrolled 21 women aged 26 to 55 across three groups for eight weeks. Its authors call it an initial study and ask for a larger sample and a longer duration 5. That is the state of head-to-head evidence in this category.
So the rubbed-on category in one line: one decent trial, run by the company that sells the ingredient, showing a peptide moisturiser beats the same moisturiser without it; one 21-person head-to-head whose own authors call it preliminary; and one independent facial trial of the copper peptide that was negative on every objective measure. That is enough to say peptide serums are not nothing. It is nowhere near enough to say any of them beats the cheaper or better-studied things they have never been tested against.
The copper peptide's one independent trial
GHK was first described in the 1970s and has been sold for skin since the 1980s, and it carries the loudest marketing in the category. Across that literature we found one independent randomised controlled trial of it on facial skin, and no other: 13 patients followed after CO2 laser resurfacing. Computer analysis and blinded evaluators found no significant improvement in redness, wrinkles or skin quality. The one measure that favoured the copper peptide was the patients' own questionnaire (P = .04) 6. An independent pharmacy-school review describes the same thing from the other direction, as a surprising absence of clinical studies 8.
GHK-Cu has one positive randomised human result. A topical gel on diabetic foot ulcers reported 98.5 percent against 60.8 percent median closure, with infection in 7 percent of patients against 34 percent 7. That was an open, limb-threatening wound managed by a clinician in 1994, the product was never approved, and it is a different situation from intact ageing skin. No copper peptide product is approved to treat a wound of any kind, and nothing in that 1994 result transfers to a serum bought off a shelf. Wound and scar context of that kind sits on our injury recovery page.
Although GHK-Cu and Pal-GHK are widely used in anti-wrinkle products, there is a surprising absence of clinical studies using them.Mortazavi, Mohammadi Vadoud and Moghimi, BioImpacts, 2025 [[8]]
That review also concludes that published information on how well these peptides cross skin, and how effective they are, is insufficient 8. The claim that GHK-Cu resets 4,048 genes does not change that: it comes from re-analysing three public microarray profiles from prostate and breast cancer cell lines with a crude cut-off and no statistical test, by authors affiliated with a company selling GHK-Cu products, and no skin cells were involved anywhere in it. Our GHK-Cu monograph sets out where that number came from.
The 55.8% wrinkle figure quoted everywhere comes from a different, smaller study that used a nano-carrier formulation and has not been independently replicated.
How much of a water-loving peptide crosses a fat-loving skin barrier at all is a formulation question rather than a percentage-on-the-label question, and how GHK-Cu works covers it, while our comparison page traces where the GHK-Cu, retinol and vitamin C headline numbers came from. GHK-Cu is marketed for hair on the same reasoning and with less published behind it than the skin claims; the hair question has its own page.
Are injectable peptides for skin worth it compared with a serum?
No randomised human skin trial exists for any injectable peptide, including BPC-157, epithalon and GHK-Cu, which is widely sold in vials even though every human study of it on skin used a cream.
A 2026 narrative review of injectable peptides, written for orthopaedic and sports medicine physicians, puts it plainly: BPC-157's reported benefits "are largely unvalidated in human trials", and GHK-Cu "showed promise in wound healing and anti-inflammatory effects, but no clinical data support its use for musculoskeletal conditions". The same review states that indications, dosing, frequency and duration remain unknown for injectable peptides generally 9. Compound detail for BPC-157, including what its evidence base really consists of, sits on the BPC-157 page, and cross-compound repair evidence sits on our healing guide.
Epithalon's headline result arrived in 2025, and it is cell culture work. Researchers at Brunel University London reported dose-dependent lengthening of telomeres, the protective caps on the ends of chromosomes that shorten as cells divide, in cultured human cell lines: breast cancer lines plus normal surface-lining cells and collagen-making cells 10 (the published literature spells the compound Epitalon). The paper's own mechanism is worth stating plainly. In the normal cells the lengthening ran through upregulation of hTERT and telomerase, and in the breast cancer lines through alternative lengthening of telomeres 10. Switching telomerase back on is one of the things malignant cells do to divide without limit, so that finding is a caution rather than a selling point, and nobody has established what dose-dependent telomerase upregulation means for a person over any timeframe. This is cell culture, not skin and not people, and the paper carries a published correction issued in November 2025. We found no human skin study of epithalon. Detail sits on the epithalon page, and the broader ageing context on our anti-ageing page.
KPV enters this conversation through acne, because it calms inflammation. It gets into cells through a transporter that lives in the gut lining, which is why almost all of its evidence is about the gut, and we found no published human trial of KPV at all, on skin or anywhere else. The KPV page has the detail and our inflammation page has the wider context. Antimicrobial peptides such as LL-37 come up for acne in the same way, with the same absence of human skin trials behind them. TB-500 appears in skin blends on soft-tissue repair reasoning, and like the rest of the injected category it has no human skin trial behind it; TB-500 has the detail.
Purity and sterility of grey-market vials is set out on the BPC-157 page, which carries the figures and their limits. It is the single most relevant fact about the injected category, and it is about the supply chain rather than about the molecule.
The two named blends sold for skin, Glow and Klow, are built from the same compounds as everything above, and no published study has tested either combination as a blend. Stacking for ageing more broadly is covered on the triple anti-ageing stack page.
Skin and appearance
GLOW
70 mg vialThe skin blend: one collagen peptide and two repair peptides
Skin and appearance
KLOW
80 mg vialGLOW with an anti-inflammatory peptide added
How does any of this compare with retinol?
Prescription tretinoin is the benchmark the whole category gets measured against, whether or not the comparison is made out loud. Its pooled record is eight randomised controlled trials and 1,361 patients: fine wrinkles, mean difference 0.412 (95 percent CI 0.233 to 0.590, p < 0.001); coarse wrinkles, mean difference 0.245 (95 percent CI 0.119 to 0.370, p < 0.001) 11. Those mean differences are movements on the graders' own wrinkle scales, and the peptide reviews use different scales, so the 0.412 here and the 0.27 in the peptide review above cannot be lined up side by side. What the tretinoin record shows is an evidence base several times larger than anything any skin peptide has. No skin peptide has been compared against it at anything like that scale, or against it at all in a published head-to-head: the category is measured against retinol constantly and has never once been put in a room with it.
The choice most readers are making
What to check before you buy, by category
The risk is different in each category, so the question worth asking before you spend anything is different too.
A serum is a concentration and formulation question
The trial that found an effect used a stated amount of peptide in a stated base: 3 parts per million of palmitoyl pentapeptide-4 in a moisturiser 4. Most cosmetic labels do not state the peptide concentration at all, which makes the product impossible to line up against any published result. For copper peptides specifically, the complex is blue when it is intact, because the colour is the copper being carried, so a preparation that has lost its colour has lost something. Storage matters for the same reason: heat, light and time all work against a peptide in solution.
A supplement is a dose and funding question
The pooled collagen reviews describe roughly 90 days of daily ingestion before measuring anything 3, so a two-week trial of a powder tells you nothing either way. Ask who paid for the study printed on the tub, because the honest answer across this literature is usually a collagen company 2. Hydrolysed collagen is made from bovine, porcine or marine sources, so it is not an option if you avoid those or react to fish, and the pooled trials ran in adults with no pregnancy data. Nobody should read a study-reported amount as a personal recommendation; what to take, and whether to take anything, is a conversation for a clinician.
An injectable is a sterility question
Before any question about whether an injected peptide works on skin, there is the question of what is in the vial and whether it is sterile. No randomised human skin trial exists for BPC-157 or epithalon 9, so the potential upside being weighed against that risk is entirely hypothetical. Injecting a non-sterile preparation risks infection at the site and worse, and injecting into the face adds blood-vessel and granuloma risks on top.
If you are comparing suppliers, our vendor comparison lists what we can verify about each. Plain disclosure, repeated here where you might act on it: those vendor links are affiliate links, and PeptidesDNA earns a commission if you buy through them. That is also why the quality data above is quoted from an independent analysis rather than from anyone selling anything.
Why the same serum helps one person and does nothing for the next
Two people buy the same serum. One sees a change at eight weeks, the other sees nothing, and their reviews read as though they are describing different products. Ranked by how much of that gap each one explains, here is what is going on.
- What was in the product. Cosmetic labels rarely state peptide concentration, and the trials that found an effect used specific amounts in specific bases. A serum carrying a trace of peptide is a moisturiser with a marketing line on the box.
- How long it was used. The trials on this page ran 8 to 12 weeks before measuring anything. Most people judge a serum in three weeks and stop.
- How it was measured. A clinical trial controls lighting, camera angle, hydration and time of day. A phone photo taken in a different bathroom at a different hour controls none of them.
- Sun exposure and the ordinary biology underneath. Skin collagen density falls steadily across adult life 12, and elastin, the protein that lets skin spring back, is barely replaced in adult skin. That is what every product on this page is being asked to push against, and our skin ageing page covers the biology and the realistic timeline.
- Genetics, last and smallest, for the reason set out below.
A photograph taken without standardised lighting, hydration state, time of day and camera distance cannot separate a change in skin from a change in those four variables, and none of the widely circulated before-and-afters standardise them.
What a before-and-after photograph can and cannot show is covered properly on its own page. If you are over 50, age-specific context sits on peptides after 50; if you are new to the category, peptides for beginners is the better starting point than this one, and what GHK-Cu is claimed to do is the page for anyone who arrived here already sold on the copper peptide.
Where genetics fits, and where it does not
We found no published study testing whether any gene variant changes how skin responds to any of the compounds on this page (literature searched September 2026). That single sentence is the honest position, and it rules out the confident story sold across peptide sites, in which your genes explain why a particular serum failed you.
The variants that get quoted do have real literature behind them, and it is literature about baseline biology rather than about response to a product. The collagen variant repeated most often is an Sp1 binding-site change in COL1A1, the gene for the main structural protein in skin (rs1800012, if you want to find it in your own raw data file); the published work on it is predominantly bone-density research, and we are not aware of a skin-collagen study of it. The collagen-breakdown marker quoted alongside it sits in MMP1, the gene for an enzyme that cuts collagen apart (rs1799750), and it is an insertion-deletion that consumer genotyping chips generally cannot read, so an uploaded raw data file usually cannot report that marker at all.
Knowing your own variants will not tell you which product to use or how much of it. It tells you how much weight to give a marketing claim built on a single headline gene, and nothing more than that. Breadth with honest confidence-grading beats one variant quoted with a confidence nobody has earned. Our decision framework works through that, the guide to reading your own raw data shows where these markers live in a file, and the connective tissue page covers the collagen variants in more depth.
Curious how your own genetics line up against the compounds on this page? Upload the raw DNA data you already have and get all 39 peptides scored for you across a 144-marker panel, 8 pharmacogenes and 15 pathways, with a human review before delivery and turnaround within 24 hours. The report grades how confident the evidence is for each compound, including everywhere it is thin.
Get your DNA report- The evidence ranking runs close to the reverse of the price ranking
The cheapest category, a collagen powder, has 26 randomised trials and 1,721 people pooled behind it. The most expensive category, injectable vials, has no randomised human skin trial at all.
- Sort skin peptides by how they get in
Swallowed, rubbed on and injected are three unrelated categories sharing one word. The evidence sits almost entirely in the first, the marketing sits almost entirely in the third, and nothing useful can be said about peptides for skin until you know which category is on the label.
- Studied is a different quantity from shown to work
GHK-Cu has a literature running back to the 1970s and one independent randomised facial trial, which was negative on every objective measure. Volume of publication and weight of evidence are not the same thing.
Sources12
- Nukaly HY, Halawani IR, Irtaza HM, et al. Oral and topical peptides for skin aging: systematic review and meta-analysis of randomized controlled trials. Frontiers in Medicine 2026;13:1618306.
- Pu SY, Huang YL, Pu CM, et al. Effects of Oral Collagen for Skin Anti-Aging: A Systematic Review and Meta-Analysis. Nutrients 2023;15(9):2080.
- de Miranda RB, Weimer P, Rossi RC. Effects of hydrolyzed collagen supplementation on skin aging: a systematic review and meta-analysis. International Journal of Dermatology 2021;60(12):1449-1461.
- Robinson LR, Fitzgerald NC, Doughty DG, et al. Topical palmitoyl pentapeptide provides improvement in photoaged human facial skin. International Journal of Cosmetic Science 2005;27(3):155-160.
- Aruan RR, Hutabarat H, Widodo AA, et al. Double-blind, Randomized Trial on the Effectiveness of Acetylhexapeptide-3 Cream and Palmitoyl Pentapeptide-4 Cream for Crow's Feet. Journal of Clinical and Aesthetic Dermatology 2023;16(2):37-43.
- Miller TR, Wagner JD, Baack BR, Eisbach KJ. Effects of topical copper tripeptide complex on CO2 laser-resurfaced skin. Archives of Facial Plastic Surgery 2006;8(4):252-259.
- Mulder GD, Patt LM, Sanders L, et al. Enhanced healing of ulcers in patients with diabetes by topical treatment with glycyl-l-histidyl-l-lysine copper. Wound Repair and Regeneration 1994;2(4):259-269.
- Mortazavi SM, Mohammadi Vadoud SA, Moghimi HR. Topically applied GHK as an anti-wrinkle peptide: Advantages, problems and prospective. BioImpacts 2025;15:30071.
- Mayfield CK, Bolia IK, Feingold CL, et al. Injectable Peptide Therapy: A Primer for Orthopaedic and Sports Medicine Physicians. American Journal of Sports Medicine 2026;54(1):223-229.
- Al-Dulaimi S, Thomas R, Matta S, Roberts T. Epitalon increases telomere length in human cell lines through telomerase upregulation or ALT activity. Biogerontology 2025;26:178 (correction published 15 November 2025).
- Huang HY, Lee LT. Tretinoin for Photodamaged Facial Skin: Systematic Review and Meta-Analysis of Randomized Controlled Trials. Dermatology Practical and Conceptual 2025;15(4):e20255172.
- Shuster S, Black MM, McVitie E. The influence of age and sex on skin thickness, skin collagen and density. British Journal of Dermatology 1975;93(6):639-643.
This article is for informational and educational purposes only. It is not medical advice and does not diagnose, treat, cure, or prevent any disease. Consult a qualified healthcare professional before starting any peptide protocol. Individual results vary. Some outbound links are affiliate links, at no extra cost to you.
Frequently asked questions
Is GHK-Cu better than Matrixyl for skin?
Neither has been shown to beat the other, because they have never been compared in a published trial. Palmitoyl pentapeptide-4, sold as Matrixyl, has the larger single trial: 93 women, 12 weeks, split-face, against the same moisturiser without the peptide, run by the ingredient manufacturer's own laboratory [[4]]. The only head-to-head trial of two cosmetic peptide creams we could find ran 21 women for eight weeks and its authors call it preliminary [[5]]. GHK-Cu has a far longer literature and one independent randomised facial trial, in 13 patients, which found no objective benefit on computer analysis or blinded assessment [[6]]. A longer literature is not a stronger result.
Does oral collagen actually work for skin?
It has the most human evidence of anything sold as a skin peptide, and the effects are modest. Two independent pooled analyses, covering 26 randomised trials and 1,721 people [[2]] and 19 randomised trials and 1,125 people [[3]], found improvements in skin water content, stretchiness and wrinkle scores against placebo after roughly 90 days. The caveats: most of those trials were funded by collagen companies, the 2023 review states that it identified several biases in the trials it pooled [[2]], 95 percent of participants in the 2021 pooled set were women [[3]], and the measurements were mostly instrument readings rather than visible change. Hydrolysed collagen comes from bovine, porcine or marine sources, so check the source if you react to fish or avoid those.
Can you use copper peptides and retinol together?
Both are commonly marketed for use in the same routine, typically at different times of day, and no published trial has tested the combination. Whether either belongs in your routine is a question for a clinician or dermatologist, not for us. Layering a copper peptide with a retinoid or an acid stacks irritation, and copper peptides on their own can cause redness, stinging, itching and flaking. Vitamin C reduces copper and is expected on redox grounds to destabilise the complex, which is why vitamin C serum and copper peptides are usually kept apart in time, though no published trial has tested that in a formulated product on skin [[8]]. Anyone with Wilson's disease or another copper-handling condition should avoid copper peptides entirely, and anyone with a cancer history, or who is pregnant or breastfeeding, should raise it with a clinician first.
What is the best peptide for acne scars?
None of them has been shown to work on acne scars in a randomised human trial. BPC-157 is the one most often suggested, and a 2026 review written for sports medicine physicians states that its reported benefits are largely unvalidated in human trials [[9]]. KPV is raised for inflammation and we found no human study of it at all, on skin or anywhere else. The interventions with published evidence for acne scarring are the dermatology ones, including laser resurfacing, clinically performed microneedling and, for raised or keloid scarring specifically, silicone gel or sheeting. Scarring is worth taking to a dermatologist rather than to a vial.
This article is for informational and educational purposes only. It is not medical advice and does not diagnose, treat, cure, or prevent any disease. Consult a qualified healthcare professional before starting any peptide protocol. Individual results vary. Some outbound links are affiliate links, at no extra cost to you.