How many grams of collagen peptides should I take per day?

It depends on what you are taking it for, and the studied doses differ a lot. Across the skin trials the dose range is roughly 2.5 to 10 grams per day for 8 to 12 weeks, with meta-analyses emphasising around 90 days; a frequently-cited skin trial tested 2.5 and 5 grams. A joint hydrolysate trial used about 10 grams per day for 24 weeks. A bone trial used 5 grams per day for 12 months. There is no established optimal dose, only doses that have been tested.

Is a lower molecular weight collagen peptide better?

The marketing implies that lower Daltons means better absorption and better results, but the human evidence does not carry it. One randomised trial compared a 5,000 Dalton and a 1,200 Dalton hydrolysate against placebo in pressure ulcer patients, but it never compared the two active arms head to head and it confounded molecular weight with dipeptide content, so molecular weight cannot be isolated as the cause. A separate human pharmacokinetic study in nine people found that ordinary collagen hydrolysate did not significantly beat plain gelatin on one key peptide measure. Treat low-Dalton claims as inadequately supported, not as a reason to pay more.

Does collagen supplementation work for hair growth?

This is the weakest claim in the category. We did not locate any placebo-controlled randomised trials of collagen peptides for hair growth. That absence is not proof that collagen does nothing for hair, but it does mean anyone selling you collagen for your hair is selling you an untested claim.

Is UC-II the same thing as collagen peptides?

No, and the two should never be blended together. UC-II is undenatured type II collagen dosed at about 40 milligrams per day, roughly 250 times smaller than the 10 grams per day used in hydrolysate joint trials. Researchers propose that it works through oral tolerance and immune modulation rather than by supplying building blocks. Evidence for one tells you nothing about the other.

Does an FDA-registered facility mean the product is proven to work?

No. Under DSHEA, the FDA does not have the authority to approve dietary supplements before they are marketed, and it generally does not approve supplement claims or labelling before use. Facility registration says nothing about whether a product works. The firm, not the FDA, is responsible for holding the substantiation behind its claims.

How independent is the collagen research?

A substantial amount of the flagship consumer evidence shares authorship. Steffen Oesser of the Collagen Research Institute in Kiel, Germany is a co-author on the flagship skin trial, the flagship nail study and the flagship bone trial. This is a descriptive observation about author lists, not an allegation of bias or misconduct, but it is a reason to want independent replication before treating any single result as settled.

Collagen & skinbest collagen peptides

Best Collagen Peptides: Choosing on the Evidence

Studied doses, third-party testing, molecular-weight marketing and cost per studied dose. The criteria that separate a good collagen from a well-advertised one.

Written bySebastian Hunter, Founder and Editor

Published July 15, 2026Updated August 22, 2026

Heads up: this page contains affiliate links, and we may earn a commission if you buy through one - at no extra cost to you. It never changes what the evidence section says. How we handle this.

Start here

Plain English

Choosing a collagen powder comes down to the dose used in the trials, the form, third-party testing and cost per studied dose. We give you the criteria rather than a single winner, because the research supports ingredients and doses rather than brands.

There is no universal best brand for everyone.
Prefer clear peptide form, study-aligned dose, and third-party testing where possible.
Marketing “types” often outrun what trials actually used.

There is no “best collagen peptide.” There are products that deliver a dose someone has actually studied, in a form you will actually take, at a price that survives arithmetic - and there is everything else. This page is the framework for telling those apart, followed by our current picks.

Moderate human evidence

Multiple randomised trials, though with limitations. Skin has 19-RCT meta-analyses behind it, but the most recent synthesis reports only a modest pooled wrinkle effect and inconsistent elasticity results. Joints rest on very low quality evidence. Bone rests on essentially one trial. Hair has no placebo-controlled trials we could locate.

US legal status - Dietary supplement

Regulated as a dietary supplement in the US. Supplements are not FDA-approved; manufacturers are responsible for their own safety and labelling, and the FDA acts only after a problem surfaces.

What actually matters when choosing

Lead with this, because it is the part that saves you money. Five criteria do almost all the work, and most product marketing addresses none of them.

1. Dose per serving, measured against the doses that were studied

This is the single most useful filter, and it is the one most often failed. The studied doses are not interchangeable, and they differ by what you are hoping for:

Skin: 2.5–10 g/day across the trials, over 8–12 weeks, with meta-analyses emphasising roughly 90 days. Proksch 2014 - Proksch E, Segger D, Degwert J et al. Oral supplementation of specific collagen peptides has beneficial effects on human skin physiology: a double-blind, placebo-controlled study. Skin Pharmacol Physiol. 2014;27(1):47-55. (opens PubMed in a new tab), a frequently-cited trial, randomised 69 women aged 35–55 to 2.5 g, 5.0 g or placebo for 8 weeks and reported a statistically significant elasticity improvement in both collagen groups versus placebo; moisture and transepidermal water loss showed positive trends that did not reach significance overall.
Joints (hydrolysate): ~10 g/day for 24 weeks - Clark 2008 - Clark KL, Sebastianelli W, Flechsenhar KR et al. 24-Week study on the use of collagen hydrolysate as a dietary supplement in athletes with activity-related joint pain. Curr Med Res Opin. 2008;24(5):1485-96. (opens PubMed in a new tab), in athletes with activity-related joint pain.
Bone: 5 g/day for 12 months - König 2018 - König D, Oesser S, Scharla S et al. Specific Collagen Peptides Improve Bone Mineral Density and Bone Markers in Postmenopausal Women-A Randomized Controlled Study. Nutrients. 2018;10(1). (opens PubMed in a new tab).
Nails: 2.5 g/day for 24 weeks - Hexsel 2017 - Hexsel D, Zague V, Schunck M et al. Oral supplementation with specific bioactive collagen peptides improves nail growth and reduces symptoms of brittle nails. J Cosmet Dermatol. 2017;16(4):520-526. (opens PubMed in a new tab), which was open-label with no control group.

A product offering 1 g per scoop is not a smaller version of the skin trials. It is outside them. Read the serving size, not the tub size.

2. Which claim you are actually buying

The evidence is not one block. It is four separate piles of very different height, and a product that markets all four is borrowing the strongest to sell the weakest.

The four claims, ranked by what supports them

Skin has the most: de 2021 - de Miranda RB, Weimer P, Rossi RC. Effects of hydrolyzed collagen supplementation on skin aging: a systematic review and meta-analysis. Int J Dermatol. 2021;60(12):1449-1461. (opens PubMed in a new tab) pooled 19 double-blind RCTs and 1,125 participants and concluded that ingestion of hydrolyzed collagen for 90 days “is effective in reducing skin aging, as it reduces wrinkles and improves skin elasticity and hydration.” The more recent synthesis is more tempered. Nukaly 2026 - Nukaly HY, Halawani IR, Irtaza HM et al. Oral and topical peptides for skin aging: systematic review and meta-analysis of randomized controlled trials. Front Med (Lausanne). 2026;13:1618306. (opens PubMed in a new tab) (19 RCTs, 1,341 participants) reported only “a modest pooled effect on wrinkle reduction (MD = 0.27, p = 0.04)” and found that “effects on elasticity and density were inconsistent,” calling for “larger RCTs with standardized outcomes and histopathologic assessment.” Elasticity was the older literature’s headline endpoint. Why the newer synthesis found it inconsistent is an open question we cannot resolve - and it governs how confidently anyone should state the skin claim.

Joints looks strong until you read the quality rating. Liu 2018 - Liu X, Machado GC, Eyles JP et al. Dietary supplements for treating osteoarthritis: a systematic review and meta-analysis. Br J Sports Med. 2018;52(3):167-175. (opens PubMed in a new tab) examined 20 supplements across 69 studies and placed collagen hydrolysate among the supplements with “large (effect size >0.80) and clinically important effects for pain reduction at short term” - but stated that “the quality of evidence was very low,” and that supplements “had no clinically important effects on pain and function at medium-term and long-term follow-ups.”

Bone rests on König 2018 - König D, Oesser S, Scharla S et al. Specific Collagen Peptides Improve Bone Mineral Density and Bone Markers in Postmenopausal Women-A Randomized Controlled Study. Nutrients. 2018;10(1). (opens PubMed in a new tab): 5 g/day versus placebo for 12 months in 131 postmenopausal women. Spine T-score changed +0.1 ± 0.26 versus −0.03 ± 0.18 (p=0.030); femoral neck +0.09 ± 0.24 versus −0.01 ± 0.19 (p=0.003). The absolute changes are small, the standard deviations exceed the mean effects, and this is essentially one trial.

Nails rests on Hexsel 2017 - Hexsel D, Zague V, Schunck M et al. Oral supplementation with specific bioactive collagen peptides improves nail growth and reduces symptoms of brittle nails. J Cosmet Dermatol. 2017;16(4):520-526. (opens PubMed in a new tab) - n=25, open-label, no control group. It reported 12% increased nail growth and 42% fewer broken nails. Without a control, placebo effect, regression to the mean and natural variation cannot be excluded.

Hair is thinnest of all. We could not locate placebo-controlled RCTs of collagen peptides for hair growth. Absence of indexed results is weaker than a positive finding, so we will not claim none exist - but nobody should be selling you collagen for hair on the strength of published trials.

UC-II is a different product - do not let a label blend them

Undenatured type II collagen (UC-II) is dosed at 40 mg/day, roughly 250 times smaller than the ~10 g/day used in hydrolysate joint trials, and researchers propose it acts through oral tolerance and immune modulation rather than as raw substrate. Lugo 2016 - Lugo JP, Saiyed ZM, Lane NE. Efficacy and tolerability of an undenatured type II collagen supplement in modulating knee osteoarthritis symptoms: a multicenter randomized, double-blind, placebo-controlled study. Nutr J. 2016;15:14. (opens PubMed in a new tab) randomised 191 subjects with knee osteoarthritis for 180 days: WOMAC total improved versus placebo (p=0.002) and versus glucosamine plus chondroitin (p=0.04). That trial was manufacturer-sponsored and authored by sponsor employees; the disclosure states it was run and managed independently by a contract organisation, with data analysis by an independent statistician. None of this transfers to a hydrolysate powder, and vice versa.

3. Third-party testing - and what it does and does not tell you

Ask for a current certificate of analysis from an independent lab, batch-matched to the tub you are buying. The honest framing: third-party testing verifies identity, purity and contaminants. It does not verify that the product does anything.

That distinction matters because of how the category is regulated. Under DSHEA, “FDA does not have the authority to approve dietary supplements before they are marketed,” and “FDA generally does not approve dietary supplement claims or other labeling before use.” Structure/function claims “are not approved by FDA and do not require FDA evaluation before they are used in dietary supplement labeling,” which is why the “not intended to diagnose, treat, cure, or prevent any disease” disclaimer is required. The firm - not the FDA - must hold the substantiation.

One gap we should be straight about: whether heavy-metal or sourcing risks differ materially by source species - BSE-prion considerations for bovine, allergen risk for marine - is something we have not investigated, and we will not hand you a ranking we have not earned.

4. Form and source - where marketing outruns the evidence

Form is mostly dose arithmetic. To reach 5 g from capsules holding a few hundred milligrams each, you are swallowing a lot of capsules, daily, for months. Powder is usually the only form where the studied doses are practical. Gummies are the worst offender: check the grams, then check how many gummies the grams imply.

Source is where we have to disappoint a popular shorthand. The common claim - marine ≈ Type I, bovine ≈ Type I+III, chicken ≈ Type II - is not supported by the review it is usually attributed to. What León-López 2019 - León-López A, Morales-Peñaloza A, Martínez-Juárez VM et al. Hydrolyzed Collagen-Sources and Applications. Molecules. 2019;24(22). (opens PubMed in a new tab) does support is tissue-level typing: Type I predominates in skin, bone, teeth, tendon, ligaments and vasculature; Type II is the cartilage collagen; Type III is common in skin, muscle and blood vessels. That review names bovine, porcine, ovine, chicken, duck, rabbit, fish and jellyfish as hydrolyzed-collagen sources, but it states no species-to-type mapping. That mapping needs a citation we could not find. Choose your source on price, dietary preference and allergen risk - not on a type-matching story.

5. Molecular weight, and the cost-per-studied-dose calculation

Lower Daltons is the category’s favourite premium. The human evidence does not carry it.

Sugihara 2018 - Sugihara F, Inoue N, Venkateswarathirukumara S. Ingestion of bioactive collagen hydrolysates enhanced pressure ulcer healing in a randomized double-blind placebo-controlled clinical study. Sci Rep. 2018;8(1):11403. (opens PubMed in a new tab) randomised 120 subjects with stage II/III pressure ulcers to a ~5,000 Da hydrolysate, a ~1,200 Da hydrolysate, or placebo. The lower-MW arm beat placebo on all three endpoints; the higher-MW arm on one. But the arms were never compared head-to-head, and molecular weight was confounded with dipeptide content - so MW cannot be isolated as the cause. It is also a wound-healing population, not a cosmetic or joint one. Separately, Iwasaki 2022 - Iwasaki Y, Nakatogawa M, Shimizu A et al. Comparison of gelatin and low-molecular weight gelatin hydrolysate ingestion on hydroxyproline (Hyp), Pro-Hyp and Hyp-Gly concentrations in human blood. Food Chem. 2022;369:130869. (opens PubMed in a new tab) (n=9) found that low-MW hydrolysate significantly raised Pro-Hyp and Hyp-Gly versus gelatin, but ordinary hydrolysate did not significantly exceed gelatin on Hyp-Gly, and those authors concluded that “gelatin is useful as a functional food as effectively as GH.”

So the low-Da premium is inadequately supported - not because nobody has run the trial, but because the trial that exists cannot separate molecular weight from what it travelled with.

We do know food-derived collagen peptides reach human blood at micromolar levels - high, relative to the nanomolar levels typical of most bioactive food peptides - and that Pro-Hyp “enhances the growth of fibroblasts and synthesis of hyaluronic acid” (Sato 2017 - Sato K. The presence of food-derived collagen peptides in human body-structure and biological activity. Food Funct. 2017;8(12):4325-4330. (opens PubMed in a new tab)). That activity is in vitro, and the same source says these observations only “partially explain” the benefits of collagen hydrolysate ingestion. Whether those peptides reach human dermis or cartilage at bioactive concentrations is not established.

So do the arithmetic instead of buying the Daltons:

cost per studied dose = (price per tub ÷ grams per tub) × grams in your target dose

Pick your target from the studied doses in criterion 1 - then compare products on that number, not on price per tub.

Our picks

Buyer’s guide: a short decision path

Name the outcome. Skin, joints, bone, nails - each has a different studied dose and a different weight of evidence. If the answer is hair, buy nothing yet.
Match the dose to the trial, not to the label’s suggested serving.
Choose powder unless the dose is small enough for capsules to be realistic - outside UC-II, it usually is not.
Ignore the molecular-weight premium until someone runs a head-to-head trial with clinical endpoints.
Ask for a batch-matched certificate of analysis.
Divide, don’t compare tubs. Cost per studied dose is the only price that means anything.
Give it 90 days - the duration the skin meta-analyses emphasise - then decide honestly whether anything changed.

What we still don’t know

We would rather say this out loud than let the roundup imply more certainty than exists.

Independent replication of the bone finding. That claim rests on essentially one trial. And the same name recurs across the category’s flagship results: Steffen Oesser of the Collagen Research Institute in Kiel co-authors the flagship skin trial (Proksch 2014 - Proksch E, Segger D, Degwert J et al. Oral supplementation of specific collagen peptides has beneficial effects on human skin physiology: a double-blind, placebo-controlled study. Skin Pharmacol Physiol. 2014;27(1):47-55. (opens PubMed in a new tab)), the flagship nail study (Hexsel 2017 - Hexsel D, Zague V, Schunck M et al. Oral supplementation with specific bioactive collagen peptides improves nail growth and reduces symptoms of brittle nails. J Cosmet Dermatol. 2017;16(4):520-526. (opens PubMed in a new tab)) and the flagship bone trial (König 2018 - König D, Oesser S, Scharla S et al. Specific Collagen Peptides Improve Bone Mineral Density and Bone Markers in Postmenopausal Women-A Randomized Controlled Study. Nutrients. 2018;10(1). (opens PubMed in a new tab)) - all three pillars of the consumer case. Schunck, also at CRI, appears on two of the three author lists. That is a description of author lists, not an accusation.
Specificity to collagen. König 2018 - König D, Oesser S, Scharla S et al. Specific Collagen Peptides Improve Bone Mineral Density and Bone Markers in Postmenopausal Women-A Randomized Controlled Study. Nutrients. 2018;10(1). (opens PubMed in a new tab) used a maltodextrin placebo - a carbohydrate, not a protein. No trial we found tests collagen against an equivalent dose of ordinary protein, so whether the effect is about collagen rather than protein is unestablished.
Whether the pooling is legitimate. Meta-analyses pool heterogeneous proprietary “specific collagen peptides” across brands, doses, sources and molecular weights. Whether that is one coherent intervention, or whether pooling inflates apparent consistency, is unresolved.
The UC-II mechanism is proposed from oral-tolerance models; we cannot confirm it is demonstrated in humans.

None of that makes collagen peptides a bad purchase. It makes them a purchase with a modest, uneven evidence base, some of it authored by people with a commercial interest in the result. Worth knowing before you spend money every month for a year.

References

Bibliographic detail is fetched from PubMed, not written by us - so a citation here cannot drift from the paper it names. Study design comes from PubMed's own tags rather than our judgement.