Do collagen peptides actually work for skin?

For hydration and elasticity, the pooled trial evidence says yes, modestly. Two meta-analyses of randomized placebo-controlled trials, 19 trials with 1,125 people (2021) and 26 trials with 1,721 (2023), found significant improvements versus placebo, and the 2021 review concluded that 90 days of supplementation reduces wrinkles and improves hydration and elasticity. The caveats are the ones the reviewers state: most trials are short, many are funded by the collagen's maker, outcomes are instrument readings rather than photographs a stranger would notice, and the 2023 review flagged several biases.

Do collagen peptides help hair growth?

The evidence is thin and recent. The largest placebo-controlled trial we could open, 114 women with self-reported thinning taking 1 g of a fish-collagen peptide daily for 24 weeks, reported a significant increase in hair diameter at week 24; it was run by the peptide's manufacturer. A second 24-week trial studied hair damage, not loss. The one trial in diagnosed hair loss used a multi-ingredient tablet alongside minoxidil or finasteride, so nothing in it can be attributed to collagen. No trial shows collagen regrows hair in androgenetic alopecia on its own.

How much collagen should you take, and for how long?

The trials in the meta-analyses used a range of doses; the 2021 review found benefit at 90 days, and the hair and nail trials ran 24 weeks. Doses in the studies on this page were 1 g per day (hair diameter), 2.5 g (nails), 3 g (hair condition) and 5 g (skin elasticity in menopausal women). There is no dose-ranging trial that identifies a best amount, and this page does not recommend one.

What are the downsides of collagen supplements?

The trials report few adverse effects; the 2026 peptide meta-analysis called them well tolerated with minimal adverse events. The downsides are elsewhere: cost over months, a product category with no pre-market efficacy requirement, source and processing that vary between brands (the 2023 review found hydration effects varied by collagen source and duration), and marketing that borrows evidence from a different product. People with fish or shellfish allergies should read the source on the label, since marine collagen is common.

Collagen Peptides for Hair and Skin: What the Trials Show, and Why a Drink Is Not a Serum

Q: Are collagen peptides the same as the peptides in skincare serums?

No. Oral collagen peptides are fragments of the collagen protein, digested in the gut to amino acids and short peptides such as glycine-proline-hydroxyproline that reach the blood; the proposed mechanism is signalling to fibroblasts from the inside. Topical signal peptides such as palmitoyl pentapeptide-4 (Matrixyl) or acetyl hexapeptide-8 (Argireline) are synthetic sequences designed to act on skin cells or nerve endings from the outside, at parts-per-million concentrations, with their own small trials. A product that cites collagen-drink trials to sell a topical peptide, or the reverse, is citing evidence for a different molecule.

Q: Is biotin better than collagen for hair?

Neither has trial evidence for hair growth in people without a deficiency. The NIH Office of Dietary Supplements says there is little scientific evidence for biotin's hair, skin and nail claims; the hair evidence is case reports in children with rare conditions, and the nail evidence is a small uncontrolled study. Collagen has the small industry trials described above. A 2026 review of nail supplements lists both as improving nail appearance and brittleness while noting that manufacturers are not required to run efficacy studies before selling.

Oral collagen peptides have two meta-analyses behind them for skin and a handful of small, mostly industry trials for hair and nails. The numbers, the funders, and the difference between a collagen you swallow and a peptide you apply.

Collagen is the best-selling "peptide" in beauty and the one with the most human trials, and both facts confuse the rest of the category. A collagen drink and a peptide serum share a word and almost nothing else: different molecules, different routes, different evidence. This page reads the collagen trials as published, separates skin from hair from nails, names who paid for the studies, and then explains why a result for a drink says nothing about a serum, or vice versa.

What an oral collagen peptide is

Hydrolyzed collagen is collagen protein, usually from fish skin or bovine hide, broken by enzymes into short chains. Swallowed, those chains are digested further; what reaches the blood is amino acids and short peptides, notably glycine-proline-hydroxyproline (Gly-Pro-Hyp), which a 2026 industry pharmacokinetic study found at roughly 54 times higher exposure after a low-molecular-weight product than after ordinary collagen (Hwang et al., Journal of Medicinal Food 2026). The proposed mechanism is that these fragments signal to fibroblasts in the dermis to make more matrix. It is a nutritional supplement, regulated as food, and manufacturers are not required to demonstrate efficacy before selling it.

Skin: two meta-analyses, one direction

Review	Trials and people	What it found	Stated limits
de Miranda et al., International Journal of Dermatology 2021	19 randomized double-blind controlled trials, 1,125 participants aged 20 to 70, 95% women	Favorable results versus placebo for skin hydration, elasticity and wrinkles; hydration and elasticity confirmed in subgroup analysis; "ingestion of hydrolyzed collagen for 90 days is effective in reducing skin aging"	Trials are short; outcomes are instrument measures
Pu et al., Nutrients 2023	26 randomized controlled trials, 1,721 participants	Significant improvement in hydration and elasticity versus placebo; hydration effects varied by collagen source and duration	"The study also identified several biases in the included RCTs"; "further large-scale randomized control trials are necessary"
Nukaly et al., Frontiers in Medicine 2026 (oral and topical peptides together)	19 randomized trials, 1,341 participants	Hydration and brightness improved; "a modest pooled effect on wrinkle reduction (MD = 0.27)", "largely driven by oral polypeptides"; elasticity and density inconsistent	Calls for larger trials with standardized outcomes and histology

That is a real, replicated, modest result for hydration and elasticity measured by instruments over about three months, in trials that are mostly small and often funded by the product's maker. It is not a result for wrinkles a stranger would see, and the 2026 review's pooled wrinkle effect is small.

Hair: small, new and industry-run

The hair evidence is where the marketing runs furthest ahead of the trials. What we could open on 2026-09-05:

Hair diameter. A randomized, placebo-controlled trial in 114 women aged 20 to 50 with thigh cellulite and self-reported hair thinning, taking 1,000 mg per day of a low-molecular-weight fish-collagen peptide or placebo for 24 weeks. Hair diameter "increased significantly at week 24, with a larger proportion of participants in the LMWCP group exhibiting measurable hair thickening." The study's affiliation is the Department of Health Food R&D at NEWTREE Co., the product's maker; hair was a secondary outcome in a cellulite trial (Hwang et al. 2026).
Hair condition, not hair loss. A randomized, placebo-controlled trial in Korean adults aged 19 to 60 "without alopecia but with mild to moderate hair damage", taking 3 g per day for 24 weeks, reported gains in gloss, tensile strength and hair diameter and no treatment-related adverse events (Kim et al., International Journal of Stem Cells 2026). It says nothing about growing hair that is not there.
Hair loss, but not collagen alone. A 12-week randomized, assessor-blinded trial in 83 people with androgenetic alopecia or chronic telogen effluvium compared a tablet containing 300 mg of marine collagen plus taurine, cysteine, methionine, iron and selenium, taken alongside the drug treatment the investigator chose, against the drug treatment alone. It was run by the product's maker, Cantabria Labs Difa Cooper. Whatever it found belongs to the tablet as a whole, with drug treatment in both arms (Milani et al., Skin Research and Technology 2023).
Menopause. A six-month randomized trial in menopausal women comparing placebo, calcium plus vitamin D, 5 g of collagen, and all three, with hair loss among the outcomes; the abstract reports skin elasticity gains (12.23% with collagen alone) and does not report a hair result (Duangjai et al., Clinics and Practice 2025).

So: one manufacturer-run trial shows thicker hair shafts over six months in women with self-reported thinning; nothing shows regrowth in diagnosed hair loss. For comparison, the finasteride label trials enrolled 1,553 men and the 5% minoxidil trial 393; our hair-peptide page sets those benchmarks out.

Nails

An open-label study of 25 women taking 2.5 g per day of a specific bioactive collagen peptide (Verisol) for 24 weeks reported a 12% increase in nail growth rate, 42% fewer broken nails, and clinical improvement in 64% of participants; there was no placebo group, and the product's maker supplied the peptide (Hexsel et al., Journal of Cosmetic Dermatology 2018). A 2026 review of nail supplements lists biotin, collagen peptides, keratin, MSM and orthosilicic acid as having "been shown to improve the clinical appearance of nails", while noting that "manufacturers are not required to carry out efficacy and safety studies before marketing" (Zaraa and Richert, Skin Appendage Disorders 2026).

The biotin comparison

Collagen is usually sold beside biotin, and the two are often in one "hair, skin and nails" product. The NIH Office of Dietary Supplements fact sheet on biotin states that supplements are promoted for hair, skin and nails "but there is little scientific evidence to support these claims"; the hair evidence it cites is case reports in children with rare hair-shaft conditions, and the nail evidence is a small uncontrolled study (NIH ODS). Neither ingredient has a placebo-controlled trial of hair growth in adults without a deficiency.

A drink is not a serum

This is the point the category blurs. An oral collagen peptide is a food-derived fragment, dosed in grams, digested, and proposed to act from the bloodstream. A topical signal peptide such as palmitoyl pentapeptide-4 (Matrixyl) or acetyl hexapeptide-8 (Argireline) is a synthetic sequence, dosed in parts per million, applied to skin, and proposed to act on cells or nerve endings it reaches through the stratum corneum. The trials for each are separate, the molecules are unrelated, and a collagen drink's meta-analysis is not evidence for a "collagen peptide" cream, which in any case cannot deliver collagen fragments through skin in the way the drink delivers them to blood. When a product page cites "clinical studies on collagen peptides" for a topical, the studies are almost always the oral trials above.

What the evidence does not show

It does not show that collagen regrows hair in androgenetic alopecia; it does not show a wrinkle change large enough to see in a photograph; it does not identify a best dose or source; and it does not show that any topical "collagen peptide" does what the oral trials found. It does show, in pooled randomized trials, modestly better instrument-measured hydration and elasticity after about three months, and, in one manufacturer's six-month trial, thicker hair shafts.

Sources and dates

PubMed searched and abstracts opened 2026-09-05: de Miranda et al. 2021 (PMID 33742704); Pu et al. 2023 (PMID 37432180); Nukaly et al. 2026 (PMID 41924746); Hwang et al. 2026 (PMID 41788055); Kim et al. 2026 (PMID 42624827); Milani et al. 2023 (PMID 37357646); Duangjai et al. 2025 (PMID 41002783); Hexsel et al. 2018 (PMID 28786550); Zaraa and Richert 2026 (PMID 40176998). The NIH ODS biotin fact sheet blocked our direct request on the day; the quoted sentence is its standing statement and is linked for verification. Corrections go to the contact page; our review standard is on the testing notes page.