A peptide serum costs somewhere between fifteen pounds and two hundred. The box makes a confident claim. And there is no way to tell from the outside whether the bottle holds a useful amount of anything.
That is a frustrating position, and not one you put yourself in. The information you would need to choose well is mostly absent from the label.
What can be said is unusually encouraging. The trials here are properly designed, the effects are small, and the risk is close to zero. Cosmetic companies run split-face studies, placebo-controlled and double-blind, in the actual target population, which beats most of what gets cited for injected peptides.
There is also a problem sitting underneath the whole category, and it concerns delivery instead of activity. The question is whether these molecules get in at all.
Can peptides actually penetrate skin?
Skin exists to keep things out, and it is good at it.
The rule of thumb in transdermal pharmacology comes from a 2000 review by Bos and Meinardi: molecules below about 500 daltons can diffuse through the stratum corneum, and essentially all topically effective drugs sit below that line. Above it, exclusion increases sharply.[^1]
There is a second barrier behind the first, and it decides what any of this is for. Wrinkles form in the dermis, the layer beneath the epidermis, where fibroblasts make collagen. So a molecule aimed at collagen production has to cross the stratum corneum, then the epidermis, and arrive somewhere deeper still in a useful amount.
Matrixyl, the most-studied cosmetic peptide, is 802 daltons. Argireline is 889. Both are comfortably over the threshold, and both are water-loving molecules trying to cross a layer built of lipids, which is a second problem on top of the first.
That does not mean nothing gets through. It means the amount that does is the central question in this category, and it is the question most product pages skip entirely.
If that feels like something you should have been able to find out before buying, it is. Molecular weight never appears on a box. Concentration usually does not either, and formulation is a trade secret. So nobody here is choosing badly, working as they are from information that has largely been withheld.
What the penetration studies found
Two findings point in opposite directions for the two best-known ingredients.
Matrixyl: the fix appears to work
The molecule sold is not plain KTTKS. It is palmitoyl-KTTKS, carrying a sixteen-carbon fatty acid tail attached specifically to help it cross a lipid barrier. A 2014 study found pal-KTTKS distributed across all skin layers while the unmodified version was undetectable, and a 2021 collaboration using mass spectrometry imaging tracked it directly.[^2]
Argireline: the measurement is unflattering
A penetration study found less than 0.2% of applied Argireline crossed the stratum corneum after 24 hours.[^3]
That figure is hard to reconcile with the marketing, because Argireline is sold on the claim that it relaxes expression lines the way botulinum toxin does. Whatever it does in a dish, under 0.2% of it arrived.
The four types of skincare peptide
They get sold as one category and they work by different routes.
| Type | Example | What it is meant to do | Evidence |
|---|---|---|---|
| Signal | Matrixyl (palmitoyl pentapeptide-4) | Tell fibroblasts to make collagen | Strongest. Split-face RCTs |
| Neurotransmitter-inhibitor | Argireline (acetyl hexapeptide-8) | Reduce muscle contraction | Weaker, and penetration is the problem |
| Carrier | GHK-Cu | Deliver copper into skin | Real trials. Our GHK-Cu guide covers it |
| Enzyme-inhibitor | Various | Slow collagen breakdown | Largely preclinical |
The signal peptides have the best case, and the neurotransmitter-inhibitor group has the most confident marketing.
Does Matrixyl work?
The anchor study for Matrixyl is a 2005 double-blind, placebo-controlled, split-face trial in 93 women, twelve weeks, one side treated and one side placebo. It is cited constantly, including in our own comparison of GHK-Cu and Matrixyl.
Sources disagree about what it found.
Several report statistically significant improvements in fine lines and wrinkles. At least one dermatology source reports the improvement as small and not reaching statistical significance.
We have not resolved that from the sources available, and it matters, because this single trial carries most of the weight in the category. Until the paper is read directly, the honest description is that the anchor trial exists, is well designed, and its result is reported inconsistently.
Two surrounding facts are less disputed. Most Matrixyl trials were funded by the ingredient manufacturer, which is ordinary in cosmetics and still shapes what gets published. And a 2005 study from Procter and Gamble in 60 women aged 35 to 65 used a different formulation over eight weeks, which is a second data point from a different company.
An independent 2023 head-to-head reported palmitoyl pentapeptide-4 outperforming both Argireline and placebo, in a small sample whose authors called for larger follow-ups.
Peptides or retinoids for wrinkles?
Most people reading about peptides for wrinkles are really weighing them against retinoids.
Retinoids are the benchmark. Decades of evidence, prescription strengths available, and effects on photoaged skin that are larger and better established than anything in this category. If the question is what has the strongest case for wrinkles, the answer is not a peptide.
What peptides offer instead is tolerability. Retinoids commonly cause dryness, peeling and irritation, particularly at the start, and a proportion of people stop using them for that reason. Peptides rarely cause any of that.
So the trade is real and it runs in the obvious direction. Stronger evidence and more side effects on one side, gentler and weaker on the other.
Plenty of people use both, and there is no established conflict between them. Whether the combination adds anything beyond what the retinoid does alone has not been tested in any way worth citing.
Are collagen supplements the same as peptide serums?
They are not, and the confusion is understandable because the names overlap.
Oral collagen peptides are swallowed. They are broken down by digestion into amino acids and short fragments, which is what happens to any protein you eat, and any effect has to survive that and then reach skin through the bloodstream.
That is a completely different proposition from a molecule applied to the surface. It has its own evidence base and its own arguments, and neither belongs here. What matters here is that the two evidence bases do not transfer in either direction.
Are peptide serums safe?
The risk is genuinely low. These are applied to intact skin at low concentrations, systemic absorption is minimal precisely because penetration is poor, and the documented harm is occasional irritation. Compare that to anything injected, which our guide on injection site reactions covers.
The claims are modest. Nobody selling a peptide serum promises what a peptide vial promises, and the endpoints in these trials are wrinkle depth and skin elasticity rather than anything that could go badly wrong.
And split-face is an elegant piece of methodology, because each participant is their own control and the two sides differ only in what was applied.
So the picture is a category with real evidence for small effects, at low risk, sold with less exaggeration than most. None of which is a recommendation, and the contrast with the rest of this subject is real.
What to look for on a peptide serum label
Concentration, if disclosed
Most products do not say. A trial at a specific percentage says nothing about a serum that keeps its own concentration private.
Palmitoyl in the ingredient name
Palmitoyl pentapeptide-4, palmitoyl tripeptide-1. That prefix is the delivery modification, and it is the difference between the molecule that penetrated in the imaging study and the one that did not.
Where the peptide sits in the list
Ingredients are listed by concentration down to 1%, after which the order is arbitrary. A peptide near the end of a long list is present in an amount nobody has to disclose, which is the thing to weigh before paying for the name on the front.
Realistic timeframes
The trials ran eight to twelve weeks. Anything promising a visible change in days is describing hydration.
How to tell whether your serum is working
The same method our guide on how long peptides take to work describes, applied to a bathroom shelf.
Judge it on the trial timescale. Eight to twelve weeks, because that is how long the studies ran and because collagen turnover is slow. Deciding in a two weeks measures hydration and nothing else.
Take a photograph first, in the same light, at the same time of day, without makeup. Skin looks different in every mirror and memory is generous. A photograph from the start is the only honest comparison available later.
Change one product at a time. Starting a peptide serum in the same week as a new cleanser and a retinoid gives you a result you cannot attribute, which is the same problem our guide on running more than one peptide describes for injections.
And expect a small change or none. That is what the trials found, measured with instruments, in the best-designed studies in this whole subject.
Where this stops being useful
Whether a specific product works, which depends on its formulation and concentration, and neither is usually disclosed.
Whether the effects are worth the price, which is a personal question with a very wide price range attached.
Whether Matrixyl's anchor trial was statistically significant, which the sources disagree about and which we have flagged rather than resolved.
Common questions
Do skincare peptides actually work?
The signal peptides have real placebo-controlled human trials behind them, and the measured effects are small. The neurotransmitter-inhibitor peptides have weaker evidence and a documented penetration problem.
Can peptides even get through skin?
The central question, in fact. Both Matrixyl and Argireline exceed the 500 dalton threshold for passive penetration. Matrixyl's palmitoyl modification appears to solve it, with imaging evidence. Argireline measured under 0.2% penetration in one study.
Is Argireline like Botox?
It is sold on that comparison. Botulinum toxin is injected into muscle. Argireline is applied to skin that keeps out more than 99% of it in the one measurement available.
Which skincare peptide has the best evidence?
Palmitoyl pentapeptide-4, sold as Matrixyl, by a clear margin. Its anchor trial is well designed, and note the disagreement about its result described above.
Are peptide serums safe?
They are among the lowest-risk things covered on this site. Applied to intact skin, minimal systemic absorption, and irritation is the usual complaint.
How long before I see anything?
The trials ran eight to twelve weeks and measured small changes with instruments. Anything visible in days is hydration rather than collagen.
Are copper peptides the same thing?
GHK-Cu is a carrier peptide and works differently. Our GHK-Cu guide covers it, including the three quite different products sold under that one name.
Peptides or retinoids for wrinkles?
Retinoids have decades of evidence and larger effects. Peptides are gentler and rarely cause the dryness and peeling retinoids can. Plenty of people use both, and whether the combination adds anything has not been properly tested.
Are collagen supplements the same as peptide serums?
No. Oral collagen is digested into amino acids before it goes anywhere, which is a different proposition entirely. A trial of one tells you nothing about the other.
How will I know if my serum is working?
Photograph your face in the same light before you start, judge it at eight to twelve weeks because that is how long the trials ran, and change one product at a time. Expect a small change or none.
Does a higher percentage work better?
One study of a 7% serum reported significant improvements, so concentration plausibly matters. Most products do not disclose theirs, which makes the question unanswerable for any particular bottle.
Sources
[^1]: Bos JD, Meinardi MM. The 500 Dalton rule for the skin penetration of chemical compounds and drugs. Experimental Dermatology 2000;9(3):165-169. Peer-reviewed review, cited consistently across the transdermal literature. Establishes that molecules below approximately 500 daltons can diffuse through the stratum corneum and that essentially all topically effective drugs and common contact allergens fall below this threshold, with exclusion increasing above it. The paper should be read at source before publication; the citation here is compiled from multiple secondary sources that agree on its content.
[^2]: Penetration of palmitoylated peptides through the stratum corneum. Reported across secondary sources. A 2014 study is described as showing pal-KTTKS distributing across all skin layers while unmodified KTTKS was undetectable. A 2021 collaboration between the University of Nottingham and a cosmetics company used 3D OrbiSIMS mass spectrometry imaging to track peptide penetration directly. Both primary publications should be located and cited before publication, since the palmitoylation argument is what makes the Matrixyl case work.
[^3]: Argireline penetration measurement. A 2015 study is reported across two independent secondary sources as finding less than 0.2% of applied Argireline crossing the stratum corneum after 24 hours. This figure is doing significant work on this page and the primary publication must be located before publishing. If it cannot be verified, the claim should be softened to describe penetration as poor without the specific percentage.
[^4]: Robinson LR, et al. Topical palmitoyl pentapeptide provides improvement in photoaged human facial skin. International Journal of Cosmetic Science, 2005. Double-blind, placebo-controlled, split-face trial in 93 women, twelve weeks. Sources disagree about the result. Several report statistically significant improvement in fine lines and wrinkles; at least one dermatology source reports the improvement as small and not statistically significant. The paper must be read directly before this page publishes, and our existing GHK-Cu versus Matrixyl comparison should be checked against whatever it says, since that page already cites this trial.
[^5]: Osborne R, et al. Split-face study of a facial moisturizer containing pal-KTTKS. Procter and Gamble, 2005. 60 women aged 35 to 65, twice daily for 8 weeks, split-face randomized design with a vehicle control. Reported across secondary sources; the publication should be linked directly. Also cited on our GHK-Cu versus Matrixyl comparison.
[^6]: Aruan RR, et al. Independent head-to-head comparison of palmitoyl pentapeptide-4 and acetyl hexapeptide-8, 2023. Reported as an independent trial finding palmitoyl pentapeptide-4 outperforming both Argireline and placebo on instrument measurement, clinical photography and self-assessment, with the authors noting a small sample size. Primary publication needed.