Peptide Glow Journal

Peptide Eye Creams: The Only Category Where Peptides Out-Trial Retinoids

Seven randomized trials have tested a peptide around the eye, against six for retinoids. What they measured, what they missed, and why dark circles are the wrong target.

Fiona G · Published 2026-09-08

Eye cream is the most expensive square centimetre in skincare, and it is usually sold on three promises at once: fewer fine lines, fewer dark circles, less puffiness. We counted the randomized trials behind peptides at that site on 2026-09-08, and the result was not the one we expected. The eye area turns out to be the only place on the face where the peptide literature is not smaller than the retinoid literature — and it is also a place where two of the three promises have never been tested at all.

The census

Search (PubMed, 2026-09-08, randomized controlled trials) Records
Peptide + periorbital / eye area / crow's feet / under-eye 7 (6 after removing a false positive)
Retinol, retinoid or tretinoin + periorbital / crow's feet / eye area 6
Periorbital hyperpigmentation or dark circles, any ingredient 21
Eye cream or periorbital, any ingredient 212

Across the face as a whole, that ratio runs the other way and not narrowly: our peptides for skin census counted 55 retinoid trials against 9 for topical peptides in wrinkles and photoaging. At the eye, six against six.

Two corrections before that number is used for anything.

One of the seven is not about skin. A 1991 paper in the Journal of Clinical Endocrinology and Metabolism on the expansion of extracellular volume and suppression of atrial natriuretic peptide after growth hormone administration was retrieved because "peptide" and "volume" both appear in it. It has nothing to do with eye creams. Counting it would have inflated the finding by a sixth, which is why census work has to be read record by record rather than reported from a number.

One record appears in both columns. The 2010 Procter & Gamble comparison enrolled 196 women with moderate to moderately severe periorbital wrinkles, and its cosmetic arm contained 5% niacinamide, peptides, antioxidants and 0.3% retinyl propionate (Fu et al., Br J Dermatol 2010). Retinyl propionate is a retinoid ester, so the trial is legitimately in both sets and is clean evidence for neither ingredient alone.

The six trials, and what each actually measured

Argireline in 60 Chinese subjects is the best-controlled single-peptide trial at this site, and the only one in the set without the manufacturer among its authors. Participants were randomized 3:1 to acetyl hexapeptide-8 or placebo, applied to periorbital wrinkles twice daily for four weeks. Subjective grading gave 48.9% total anti-wrinkle efficacy against 0% under placebo; objective analysis of silicone skin replicas showed all roughness parameters reduced in the treatment group (p<0.01) with no clear change under placebo (Wang et al., Am J Clin Dermatol 2013). Four weeks is short and 60 people is small, but the design is a real randomized placebo comparison with an instrument endpoint.

OS-01 (Pep 14) was tested in a 12-week split-face, double-blind, vehicle-controlled study in 22 participants. The peptide formulation significantly reduced transepidermal water loss against both baseline and the vehicle, and expert grading with Antera 3D imaging showed reduced wrinkle appearance and indentation in the periorbital area. Texture and radiance improved on both sides of the face, with the peptide side better (Zonari et al., J Cosmet Dermatol 2024). All the first-listed authors are at OneSkin Inc., which sells it. The split-face design is the strongest feature here: each participant is their own control.

The 2010 regimen comparison found the cosmetic arm improved wrinkle appearance significantly more than 0.02% tretinoin at 8 weeks and comparably at 24 weeks, with significantly better tolerance by every measure through 8 weeks. Only 25-person cohorts continued to 24 weeks. Since the cosmetic arm contained a retinoid ester, the result reads most defensibly as a tolerance finding rather than a peptide finding.

The remaining three — a 2026 study combining a recombinant type III collagen injection with a collagen-III multi-peptide serum, a 2025 twelve-week trial of porcine placenta peptides taken as a functional food, and a 2010 study of a triple peptide complex — each measure something other than a peptide cream applied to an eye. An injection plus a serum cannot separate the two. An oral supplement is not a topical. The triple-complex study reports on a combination.

The pattern in the author lists

Four of the six have the product's maker among the authors. That is disclosed in each case, and it is not misconduct; it is what cosmetic ingredient research is. But it is rarely mentioned to the person reading a product page, and it is checkable in thirty seconds from the affiliations on any of these papers. The Argireline trial, run from the Second Hospital of Xi'an Jiaotong University, is the one in this set without that structure, which is a reason to weight it more heavily than its four-week duration would otherwise justify.

The two promises nobody has tested

Dark circles. PubMed holds 288 records on periorbital hyperpigmentation and dark circles, 21 of them randomized trials. None of the peptide eye trials used dark circles as an endpoint. The gap matters because dark circles are not one condition. Pigment sitting in the skin, blood vessels showing through unusually thin skin, and shadow cast by the tear trough and orbital rim produce the same appearance by three different routes, and only the first is a surface problem a topical could plausibly reach. A cream with randomized evidence for crow's feet has produced no evidence about any of the three.

Puffiness and bags. These involve orbital fat, the retaining septum and overnight fluid distribution — structures beneath the skin entirely. No trial in the set used them as an endpoint.

So the category's evidence covers one of its three headline promises. That is worth more than nothing, and considerably less than the packaging implies.

Why the eye area is a real target anyway

Eyelid skin is the thinnest on the body, and the area is in near-constant motion, which is why expression lines appear there first and why it is the site where retinoid irritation is least tolerable. That combination explains the census result: the eye is precisely where a better-tolerated but weaker option has the most room to be studied, and where the strongest ingredient is hardest to use.

It does not, however, justify the separate product. No trial has compared an eye-specific formulation against the same brand's facial product applied to the same area. The site is a legitimate target; the separate jar is an untested proposition.

The penetration arithmetic applies here as everywhere: acetyl hexapeptide-8, the peptide with the best eye-area evidence, is 887.0 daltons, well above the roughly 500-dalton threshold for passive movement through intact stratum corneum. Our peptides for skin page carries the full measured table and the honest counterweight — molecular weight sets the difficulty of the problem rather than settling it. Thinner skin at the eyelid may help; nobody has measured whether it does.

What would settle it

A split-face trial of a single peptide at a stated concentration against its own vehicle, with dark circles graded by cause rather than by appearance, run past twelve weeks by investigators without a commercial interest. Six trials in, none of that description exists.

More from this publication: the whole category counted in peptides for skin, ingredients ranked by evidence in best peptides for skin, the ingredient with the eye-area evidence in Argireline, and the same treatment of a neighbouring category in peptide lip treatments. Our testing notes explain how we handle claims that outrun their evidence.

Sources

  • PubMed searches run 2026-09-08 via NCBI E-utilities esearch, title/abstract terms as described in the table; publication type filter "randomized controlled trial"[pt]. Each record in the peptide set was retrieved and read; the 1991 endocrinology paper was excluded on inspection. PubMed.
  • Wang Y, Wang M, Xiao S, Pan P, Li P, Huo J. The anti-wrinkle efficacy of argireline, a synthetic hexapeptide, in Chinese subjects: a randomized, placebo-controlled study. Am J Clin Dermatol 2013;14(2):147-53. PubMed 23417317.
  • Zonari A, Brace LE, Harder NHO, Harker C, Oliveira CR, Boroni M, Carvalho JL. Double-blind, vehicle-controlled clinical investigation of peptide OS-01 for skin rejuvenation. J Cosmet Dermatol 2024;23(6):2135-2144. PubMed 38400612. Authors at OneSkin Inc.
  • Fu JJ, Hillebrand GG, Raleigh P, et al. A randomized, controlled comparative study of the wrinkle reduction benefits of a cosmetic niacinamide/peptide/retinyl propionate product regimen vs. a prescription 0.02% tretinoin product regimen. Br J Dermatol 2010;162(3):647-54. PubMed 20374604. Authors at The Procter & Gamble Company.

Published by Peptide Glow Journal Editorial. No product is ranked, scored or affiliate-linked on this page.

Frequently asked questions

Do peptide eye creams work?

For the appearance of expression lines at the outer eye, there is more randomized evidence than most people assume — six genuine trials, which is slightly more than retinoids have at that site. The best of them tested Argireline against placebo in 60 people for four weeks and found 48.9% anti-wrinkle efficacy against 0% for placebo, with instrument-measured roughness reductions. What none of them tested is dark circles or under-eye bags. So the category has real, modest evidence for one of the three things people buy it for.

Do peptides help dark circles?

No trial has tested that. On 2026-09-08 PubMed held 288 records on periorbital hyperpigmentation and dark circles, 21 of them randomized trials, and not one of the peptide eye-area trials used dark circles as an endpoint. This matters more than a missing study usually does, because dark circles have several distinct causes — pigment in the skin, visible blood vessels through thin skin, and shadow cast by the bony orbit and tear-trough anatomy. Only the first is a skin-surface problem at all. A topical that improves wrinkle appearance is not addressing the other two.

Are peptides better than retinol for the eye area?

At that specific site the evidence bases are unusually close: six randomized trials for peptides against six for retinoids, where across the face as a whole retinoids lead by a wide margin. Two qualifications keep that from being a win. One trial appears in both counts, because its cosmetic arm contained a retinoid ester alongside peptides. And peptide eye trials are more often run by the ingredient's maker. What the comparison does support is the tolerance point: the eye area is where retinoid irritation is worst, and it is exactly where the better-tolerated option has managed to accumulate a comparable number of studies.

Why is eye cream sold separately from face cream?

Partly anatomy and partly commerce, and the honest answer includes both. Eyelid skin is the thinnest on the body, the area moves constantly, and it is where expression lines appear first — so the outer-eye wrinkle is a genuinely distinct target, and it is the endpoint the trials use. But no trial has compared an eye-specific formulation against the same brand's face product applied to the eye area. The separate category is well founded as a target and untested as a product decision.

What concentration of peptide is in an eye cream?

Almost never stated, and the numbers that are stated usually describe a supplier solution rather than the peptide. Cosmetic peptides are sold to formulators as dilute solutions, so a label reading 10% of a named complex can contain a few thousandths of a per cent of actual peptide. The working concentrations in the trials are correspondingly small. Because no dose-response curve has been published for any cosmetic peptide, a larger label number cannot be read as a larger effect.

How long do peptide eye creams take to work?

The trials that exist ran 4 weeks (Argireline), 8 weeks (the 2010 regimen comparison at its first endpoint), 12 weeks (the OS-01 split-face study) and 56 days elsewhere in the peptide literature. Almost nothing in this field runs past 12 weeks. That is a limit on what is known rather than a schedule: no trial has followed a cosmetic peptide long enough to say whether an effect at 12 weeks continues, plateaus or reverses.

Do peptide eye creams help under-eye bags?

Nothing in the literature supports that. Under-eye bags are principally a matter of orbital fat, the septum that holds it and fluid distribution — structures well below where any topical acts, and the reason the surgical and injectable literature for that complaint is entirely separate. None of the six peptide eye-area trials used bags or puffiness as a primary endpoint. A cosmetic that improves the appearance of fine lines is not making a claim about anatomy underneath.