Best Peptide Serums of 2026: Nine Compared on Ingredients and Evidence

A peptide serum is a light treatment step built around short chains of amino acids that are designed to signal skin to make collagen, soften expression lines or carry copper. The nine serums here use very different peptides, and the human evidence behind them ranges from a 93-person placebo-controlled trial¹ to no topical human trial at all². Only one of the nine states a percentage for every peptide it contains, three headline a percentage without saying what it measures, and two sell a "copper peptide" that is not GHK-Cu. We read every label, ingredient list, shipping page and returns policy on the brand's own US site on 4 November 2026, and matched each peptide to its published studies.

Best for a stated GHK-Cu dose with every peptide disclosed: koppā GHK-Cu 2% High Potency Serum (our own)

Best budget multi-peptide serum: The Ordinary Multi-Peptide + HA Serum

Cheapest bottle, for expression lines: The Ordinary Argireline Solution 10%

Best budget GHK-Cu blend with a percentage on the label: The Ordinary Multi-Peptide + Copper Peptides 1% Serum

Most peptides named, and certified vegan: Medik8 Liquid Peptides

Best mid-price serum to layer with retinol: Naturium Multi-Peptide Advanced Serum

Lowest price per ml, and tested on sensitive skin: COSRX The 6 Peptide Skin Booster Serum

Best-studied single signal peptide: Olay Collagen Peptide Visibly Firming Serum

Newer peptides, with a refund on used product: Paula's Choice Pro-Collagen Multi-Peptide Booster

koppā makes copper peptide skincare, so GHK-Cu is the ingredient we know best, and we make the first serum on this page. That is a conflict of interest, and we handle it by showing our working: every product figure comes from the brand's own product, shipping and returns pages, read on 4 November 2026 and linked from each card, and every study is described by what kind of study it was, who ran or paid for it and how many people took part.

No brand paid to be included, and links to other brands earn us nothing. koppā wins four of the fourteen categories in our scorecard and loses the rest, including price per ml, largest bottle, most peptides named, finished-product testing, returns and vegan. Where the research is thin, we say so: a systematic review found only 6 of 15 studies of collagen-signalling peptides were placebo-controlled³, and most of the trials on this page were run or funded by the company that makes the ingredient or the product.

Three things, and they matter more than any single product on this page.

"Copper peptide" does not always mean GHK-Cu. Medik8 and Naturium both sell their serums on copper peptides, and both list Copper Palmitoyl Heptapeptide-14, a different molecule from GHK-Cu (Copper Tripeptide-1). The GHK-Cu research on this page⁹¹⁰ does not transfer to it, and we found no human trial of it on PubMed. Of the nine, GHK-Cu is listed by koppā, The Ordinary's copper serum and COSRX.

Medik8's "30% blend" is very unlikely to be 30% peptide. This is our inference from its ingredient list, not something Medik8 states: every one of its peptides is listed after phenoxyethanol, a preservative used at up to 1%, so on the usual ingredient-order rules each peptide is at or below about 1% of the formula. The 30% most probably describes peptide-containing ingredient solutions as supplied. Medik8 does not say what the figure measures.

No other brand says what its headline percentage measures. The Ordinary's "10%" Argireline and "1%" copper peptides and Medik8's "30%" do not say whether they mean pure peptide, a supplier solution or a blend. koppā states its 2% as a percentage of the finished formula by weight, though its smaller supporting-peptide percentages carry the same ambiguity. The other five serums state no peptide percentage at all.

Published trials have combined peptides with both without reported problems; the cautions that exist are brand advice, mostly about copper peptides. Procter & Gamble's randomised trial in 196 women used a regimen of niacinamide, two palmitoyl peptides (including Palmitoyl Pentapeptide-4), carnosine and 0.3% retinyl propionate, with a vitamin C derivative in the day lotion: at 8 weeks it improved wrinkles more than 0.02% prescription tretinoin, was comparable at 24 weeks, and was better tolerated²⁵. P&G funded it, five authors worked for P&G and six were paid consultants. The Boots trial put the Matrixyl 3000 pair, retinyl palmitate and a vitamin C derivative in one serum for 12 months¹⁶, and an open study put 15% L-ascorbic acid and Palmitoyl Tripeptide-38 in one serum for 56 days¹⁸.

We found no human trial showing that retinoids or vitamin C inactivate peptides. The Ordinary advises against using its copper serum with retinoids or direct vitamin C, and its other two peptide serums with direct vitamin C or direct acids. For GHK-Cu there is a lab reason for caution, not proof: it broke down under alkaline and oxidising stress, and less under acid, though it stayed stable from pH 4.5 to 7.4 for at least two weeks at 60 °C⁶. koppā's directions keep low-pH vitamin C and exfoliating acids in a separate routine and retinoids at the other end of the day. Naturium and COSRX say their serums can be layered with retinol, and Medik8 pairs its serum with vitamin C in the morning and vitamin A at night.

Quite a lot, and it is the reason we hedge every verdict on this page.

The trials are small and short. Most of the peptide trials here enrolled 13 to 93 people, often 20 to 60, for 4 to 12 weeks, and many enrolled only women¹¹⁵⁵.

Few are properly controlled. A 2019 systematic review of 15 studies of collagen-signalling peptides found only 6 were placebo-controlled, 5 double-blind and 9 in women only, and called the literature "sparse"³.

Most were run or paid for by the maker. Procter & Gamble, Boots, Sederma, Snowberry, L'Oréal and Alumier Labs ran or funded trials of their own peptides or products^1161771918. The independent FDA penetration study is the exception, and its result was the least flattering⁸.

Peptides are rarely tested alone. The Boots, P&G, L'Oréal and Alumier formulas combined peptides with retinoid esters, niacinamide or vitamin C, so the peptide's own contribution cannot be separated²⁵.

Delivery to the deeper skin is unproven. No study here shows a cosmetic peptide reaching the dermis or muscle in people²⁴.

Doses are rarely stated and never compared. Trials used 3 to 5 ppm of Palmitoyl Pentapeptide-4 and Palmitoyl Tripeptide-38 and 10% Argireline; we found no human dose-response data, and only koppā states a percentage for every peptide in its serum. The serum brands' own finished-product tests are smaller still, and none reports a control group.

A peptide serum suits you if you want a gentle treatment step for the look of fine lines and firmness, or something better tolerated than a retinoid. Look elsewhere, or at a different serum on this page, if you:

want the lowest price per ml: COSRX (US$0.17) or The Ordinary Multi-Peptide + HA (US$0.56 to US$0.66);

want to try one expression-line peptide cheaply: The Ordinary Argireline Solution 10% (US$9.70);

want GHK-Cu in a multi-peptide blend under US$35: The Ordinary Multi-Peptide + Copper Peptides 1% (US$32), accepting the unexplained 1%;

want the most peptides, or a certified vegan formula: Medik8;

may want to return an opened bottle: The Ordinary (365 days, half full), Paula's Choice (60 days, used) or Medik8 (30 days, half full);

want a finished serum with any reported test: Medik8, The Ordinary Multi-Peptide + HA or COSRX;

have a copper allergy or Wilson's disease: choose a serum without a copper peptide, such as The Ordinary Multi-Peptide + HA, The Ordinary Argireline, Olay or Paula's Choice;

prefer silicone-free: skip Naturium, Olay and Paula's Choice.