Peptides in skincare: what the evidence supports and what it does not

The honest summary first. Peptides are well tolerated, and that is the clearest thing the research shows. The evidence that a peptide cream changes how your skin looks is thinner than the marketing suggests. The large 2026 review that gets quoted for peptides is mostly a review of oral supplements rather than serums. And no peptide cream does what an injection does.

What a peptide actually is

Peptides are short chains of amino acids, the same building blocks that make up proteins such as collagen and elastin. The idea behind putting them in skincare is that they act as messengers, giving cells a nudge to behave a certain way.

The theory is reasonable. Skin does use peptide signalling. The question is how much of that survives being put in a jar, rubbed on a face, and asked to travel through a barrier whose whole job is keeping things out.

The four kinds you will see on a label

These groupings describe how each type is sold. They are not evidence that any of it happens in your skin.

• Signal peptides, such as palmitoyl pentapeptide-4, sold as Matrixyl. Marketed on the idea that they encourage the skin's own collagen.
• Carrier peptides, such as copper peptides (GHK-Cu). Marketed on the idea that they carry a trace mineral to where it is useful.
• Neurotransmitter-inhibiting peptides, such as acetyl hexapeptide-8, sold as Argireline. Marketed on the idea that they soften the look of expression lines.
• Enzyme-inhibiting peptides. Marketed on the idea that they slow the breakdown of the collagen and elastin already in the skin.

The big 2026 review is mostly about supplements

You will see a 2026 systematic review and meta-analysis cited as proof that peptides work. It pooled 19 randomised controlled trials and 1,341 people (Nukaly et al., 2026). It is a good piece of work, and the authors declared no commercial funding, which is worth something in this field.

Read the detail and it is not really about creams. Seventeen of the 19 trials tested oral peptides. Two tested topical ones. Of the 1,341 participants, 1,236 took an oral supplement and 105 used something applied to the skin.

That imbalance runs through every result. The improvements in hydration and brightness were driven by oral formulations. The pooled effect on wrinkles was small and only just reached statistical significance, and the authors report that oral polypeptides drove most of it, while the topical group showed a smaller effect that was not statistically significant. Effects on elasticity were not significant at all. For skin density, the authors note the effect was significant when peptides were taken orally, while topical use had minimal impact.

The authors say it plainly in their limitations. Only two topical trials met their inclusion criteria, and the findings primarily reflect oral peptide efficacy. So when a brand cites this review on a serum box, it is borrowing the credibility of research into supplements. That is the single most useful thing to know about peptide marketing right now.

The actual evidence for creams

Two trials carry most of the weight, and both deserve their caveats stated out loud.

The most cited one is also the oldest. Over 12 weeks, 93 women aged 35 to 55 used a plain moisturiser on one side of the face and the same moisturiser containing palmitoyl pentapeptide on the other (Robinson et al., 2005). The peptide side did measurably better for the appearance of fine lines and wrinkles, on both image analysis and expert grading, and the peptide was well tolerated. The design was genuinely good: split-face, double-blind, randomised left to right.

It was also run by researchers at Procter and Gamble, a company that sells skincare containing that peptide. That does not make it wrong. It does mean it should not be the only thing you weigh. One detail rarely mentioned: the peptide was used at 3 parts per million, which is 0.0003%. The best topical result in this field came from a very small amount of active.

The second trial tested acetyl hexapeptide-8 around the eyes. Sixty people were randomised three to one, so about 45 used the peptide and about 15 used a placebo, twice daily for four weeks (Wang et al., 2013). The researchers reported an efficacy rate of 48.9% in the peptide group against 0% in the placebo group on a subjective grading scale, and surface roughness measured from silicone skin replicas went down.

Take that with the caveats it earns. It is four weeks. The placebo group was about fifteen people. The headline outcome is a grading scale, not an objective measure. No funding source is stated either way. And this trial is one of the two topical studies inside the 2026 review above, so it is not separate confirmation of it. It is the same evidence counted once.

Two reasonable trials, on two different peptides, over twenty years. That is a thin base for a category this large. Both measured how skin looked, which is the right question for a cosmetic, and neither shows anything about treating a skin condition.

What these ingredients cannot do

They cannot do what an injectable does. Not a milder version, not a slower version.

The paper that started the comparison says so itself. In the original 2002 work, an emulsion containing 10% acetyl hexapeptide-8 reduced measured wrinkle depth by up to 30% over 30 days in healthy women volunteers (Blanes-Mira et al., 2002). Two things about that. The published account of the human test does not report how many volunteers took part or describe a placebo comparison. And the paper was written by the team that designed and named the peptide.

The same paper reports laboratory work showing the peptide interfered with the protein complex involved in neurotransmitter release, and the authors were explicit that it did so with much lower efficacy than the neurotoxin. That mechanism was demonstrated in the laboratory. It is not the same thing as showing the mechanism happened in the volunteers' faces.

Then there is delivery, which settles most of the argument. Researchers applied a 10% acetyl hexapeptide-8 emulsion to donor human skin mounted in laboratory diffusion cells and measured where it had gone after 24 hours (Kraeling et al., 2015). Most of it washed off the surface. Of the applied dose, 0.22% sat in the stratum corneum, the top layer of dead cells. Around 0.01% was found in the epidermis. None was detected in the dermis, or in the fluid beneath the skin.

That is one formulation on donor skin rather than living skin, so hold it loosely. But facial muscles sit well below the dermis, and a molecule that cannot be detected in the dermis is not plausibly reaching them.

A 2025 review reached the same place in plainer words. Acetyl hexapeptide-8 is water-loving and relatively large, its permeability through the skin barrier is limited, and its ability to reach neuromuscular junctions "remains uncertain" (Zdrada-Nowak et al., 2025). That review accepts studies suggesting it may reduce wrinkle depth and improve the look of hydration and elasticity, while stating that the mechanism behind those effects after topical use remains incompletely understood.

So do not expect any change in how your face moves. No topical peptide has been shown to do that.

Copper peptides deserve a separate note

Copper peptides have the most interesting laboratory story and the thinnest human one. A 2025 review states that on the basis of cellular studies, GHK can be considered an anti-wrinkle ingredient (Mortazavi et al., 2025). Cell studies are not faces, and the same review goes on to report a surprising absence of clinical studies using GHK-Cu and Pal-GHK, despite both being sold widely. It also notes formulation challenges, largely because these peptides are water-loving and unstable.

To be fair to the ingredient, that review does not accuse it of poor skin penetration. It concludes the opposite, describing these derivatives as relatively skin permeable. The gap is not delivery. The gap is that almost nobody has run the human trials.

None of that is a reason to avoid them. If you want one, Medik8 Copper PCA Peptides is the copper peptide serum we stock. Buy it knowing the human evidence has not caught up with the laboratory headlines.

A word about concentration

The Expert Panel for Cosmetic Ingredient Safety concluded in 2025 that acetyl hexapeptide-8 amide is safe in cosmetics in current practices of use at concentrations up to 0.005%, and that available data are insufficient to conclude it is safe above that (Johnson et al., 2025). Efficacy and penetration research has used 10%. That is a gap in the safety data rather than evidence of harm, and it is a United States panel rather than New Zealand law, but it is worth knowing when a label boasts about a big number.

It also cuts the other way. The Robinson trial got its result at 0.0003%, so a low concentration on a label is not automatically a sign of a weak product.

Using peptides sensibly

Wear sunscreen every day. This is not a throwaway line. In a randomised community trial in Queensland, 903 adults were assigned to daily broad-spectrum sunscreen or to using it at their own discretion. After four and a half years the daily group showed no detectable increase in skin ageing, and 24% less than the discretionary group (Hughes et al., 2013). It was publicly funded. That is stronger evidence than anything in the peptide literature, from a climate much like ours.

Beyond that, the usual sensible things. Patch test a new product on a small area first. Introduce one new product at a time, so you know what is doing what. Stop if your skin becomes sore, itchy or red, and give it time to settle. Do not apply anything to broken skin, and keep serums out of the eye itself rather than on the lid margin.

If you are considering an oral collagen supplement on the back of the research above, talk to your GP or pharmacist first, particularly if you are pregnant, breastfeeding, taking other medicines or managing allergies. Supplement sources vary, including marine and animal-derived ones, so check the label if that matters to you.

How to judge a peptide product fairly

Almost every trial in this field tests a finished product, not a lone peptide. That product also contains moisturising ingredients, and often other actives. Some of what people see is a well-formulated moisturiser doing exactly what a moisturiser does. That is fine. It is just not the same claim. A cream such as Medik8 Advanced Pro-Collagen Peptide Cream can leave skin looking smoother for reasons that have as much to do with the base as the peptide.

Where peptides genuinely earn their place is tolerance. Across the pooled trials, participants generally got on well with them whether taken orally or applied to the skin, with minimal adverse events reported, and the Robinson trial specifically noted good skin tolerance. If retinoids or acids leave your skin sore, something like Medik8 Liquid Peptides is an easier thing to wear daily. That is a real practical advantage, and it is a more honest reason to buy one than any wrinkle promise.

A fair expectation is this. A well-formulated peptide serum can support the appearance of smoother, better-hydrated skin. The change will be modest, and part of it will be the formula rather than the peptide. The topical trials ran between four and twelve weeks, so judge yours over a couple of months rather than a fortnight, and keep the sunscreen on throughout.

When to see a doctor instead

None of this evidence supports using a peptide product for a skin condition. Acne, rosacea, eczema, psoriasis, melasma and persistent pigment changes are medical conditions. Please see your GP or a dermatologist about them. A cosmetic serum is not an alternative to medical care, and we would never suggest you delay seeing someone.

A new mole, a spot that is changing shape or colour, or anything that bleeds or will not settle is a matter for your doctor rather than a skincare routine. Skin cancer is common in New Zealand, and that is worth taking seriously.

If you would like help working out whether peptides belong in your routine, come and talk to us in Parnell. We would rather tell you what something realistically does than sell you a promise.