supplements

How Much Collagen Per Day? What the Trials Actually Used

Collagenium
Collagenium

Editorial & Research Team

Published Last reviewed 14 min read
Scoop of collagen powder being measured into a mug

This content is for informational purposes only and is not medical advice. Collagenium is published by a research-driven editorial team, not licensed medical professionals. See our editorial standards and methodology. Consult a registered dietitian, physician, or pharmacist before starting any supplement, especially if you have existing health conditions, are pregnant or breastfeeding, or take medications.

There is no official answer to how much collagen per day you should take, because no regulator has set a recommended intake or an upper limit. What exists is a set of clinical trials, and they used fairly consistent amounts: 2.5 to 10 grams of hydrolyzed collagen a day for skin, 5 to 10 grams (or 40 milligrams of undenatured type II) for joints, 5 grams for bone density, 2.5 grams for nails and 15 grams for muscle alongside resistance training. Below we lay out those trials one by one, with participant numbers, duration and sponsor, so you can see where each figure comes from.

The short answer: doses by goal, from the trials

Every figure in this table comes from a trial we opened and checked. Doses are per day. “n” is the number of participants. The evidence label is ours and follows our methodology: well-supported means several placebo-controlled trials plus at least one meta-analysis point the same way; some evidence means one or two placebo-controlled trials; preliminary means uncontrolled or very small work.

GoalDaily dose usednDurationDesignProduct and sponsor noteEvidence
Skin elasticity and hydration2.5 g or 5 g69 women, 35-558 weeksPlacebo-controlled RCT (Proksch, 2014)Branded bovine peptide; co-author at the Collagen Research InstituteWell-supported (with the meta-analyses below)
Eye wrinkles2.5 g114 women, 45-658 weeksPlacebo-controlled RCT (Proksch, 2014)Verisol; same research groupSome evidence (one trial; wrinkles also improved in the pooled 2021 analysis)
Skin hydration, elasticity, roughness, density2.5 g plus vitamins and zinc72 women, 35+12 weeksPlacebo-controlled RCT (Bolke, 2019)Multi-ingredient drink (Elasten); run by a contract lab, sponsor-fundedSome evidence for this product
Skin (pooled)2.5 to 10 g1,125 across 19 RCTsmostly 8-12 weeksMeta-analysis (de Miranda, 2021); 26 RCTs, 1,721 people (Pu, 2023)Mixed products; authors note bias in several trialsWell-supported for hydration and elasticity
Activity-related knee discomfort5 g139 active adults12 weeksPlacebo-controlled RCT (Zdzieblik, 2017)Branded peptide; Collagen Research Institute co-authorSome evidence
Joint pain in athletes10 g147 randomized, 97 analysed24 weeksPlacebo-controlled RCT (Clark, 2008)CH-Alpha liquid; manufacturer co-authorSome evidence
Knee osteoarthritis symptom scores40 mg undenatured type II191 adults180 daysPlacebo- and comparator-controlled RCT (Lugo, 2016)UC-II; two authors employed by the makerSome evidence
Knee function after exercise40 mg undenatured type II55 healthy adults120 daysPlacebo-controlled RCT (Lugo, 2013)UC-II; maker-affiliated authorsSome evidence
Bone mineral density5 g131 postmenopausal women12 monthsPlacebo-controlled RCT (Konig, 2018)Branded peptide; Collagen Research Institute co-authorSome evidence (single trial)
Brittle nails2.5 g25 adults24 weeksOpen-label, no placebo (Hexsel, 2017)VerisolPreliminary
Muscle mass and strength15 g53 older men; 57 young men; 77 women12 weeks, each with resistance training 3x a weekPlacebo-controlled RCTs (Zdzieblik, 2015; Kirmse, 2019; Jendricke, 2019)Branded peptides in two of three; same German group in twoSome evidence, only with training

Two things stand out. First, the beauty outcomes used the smallest doses: 2.5 grams was enough to beat placebo for skin elasticity and wrinkles in the 2014 trials, and giving 5 grams instead of 2.5 did not produce a measurably better result. Second, the numbers people quote for joints, bone and muscle are all in the 5 to 15 gram range, and no placebo-controlled trial in this set used more than 15 grams a day. Claims of a 20 to 40 gram “optimal” range do not come from these trials.

No. Collagen has no Recommended Dietary Allowance, no Adequate Intake and no Tolerable Upper Intake Level, because it is not an essential nutrient: your body makes its own from amino acids in ordinary protein. It is sold as a dietary supplement, and the FDA states that it does not approve dietary supplements or their labels before they are sold; the manufacturer decides the serving size. That is why one tub says 10 grams, another says 20 and a bottle of tablets says 6,000 milligrams, and none of them is “the” dose.

Researchers say the same thing about collagen dosage from the other direction. The 2019 systematic review of collagen trials in dermatology (Choi et al.) concluded that further studies are still needed “to determine optimal dosing regimens.” In practice, the only grounded way to pick an amount is to look at what the trials for your goal actually used, which is what the table above does.

What the milligrams on a label really mean

Two of the most-searched questions on this topic are “Can I take 10,000 mg of collagen a day?” and “Is 2,000 mg enough?” Both are unit questions in disguise. One gram is 1,000 milligrams, so:

Label figureIn gramsCompared with the trials
1,000 mg1 gBelow every placebo-controlled hydrolyzed collagen trial in our table
2,000 mg2 gBelow the 2.5 g skin minimum; far below joint and bone doses
2,500 mg2.5 gThe lowest dose that beat placebo for skin and the dose in the nail study
6,000 mg6 gInside the skin range; just above the 5 g knee and bone trials
10,000 mg10 gThe top of the skin range and the dose in the 24-week athlete joint trial
20,000 mg20 gAbove anything tested in a placebo-controlled trial in our set

So “10,000 mg” is simply 10 grams, a normal trial-range dose written in a unit that makes it look large. Treat the milligram presentation as marketing rather than a sign of potency. And 2,000 mg is 2 grams, which sits under the lowest effective skin dose and well under the joint and bone doses. The one place a tiny number is legitimate is undenatured type II collagen, which was tested at 40 milligrams; it works differently, as the joints section explains, so you cannot compare it gram for gram with peptides.

Our product catalogue shows how this plays out. Youtheory Advanced Collagen lists 6,000 mg per daily serving, which is 6 grams; its label describes 290 tablets as a 48-day supply, so that serving is about six tablets. Horbaach Multi Collagen capsules list 2,000 mg per serving, so one serving is 2 grams. Double Wood liquid collagen lists 2,500 mg per serving, which is 2.5 grams, the lowest dose that beat placebo for skin.

How much collagen for skin?

The placebo-controlled skin trials used 2.5 to 10 grams of hydrolyzed collagen a day for 8 to 12 weeks, and the smallest dose in that range already produced measurable results. In the 2014 Proksch trial, 69 women aged 35 to 55 took 2.5 grams, 5 grams or placebo for 8 weeks; both collagen groups improved skin elasticity compared with placebo, and the 5 gram group did not do better than the 2.5 gram group. A companion 2014 trial in 114 women aged 45 to 65 gave 2.5 grams of Verisol or placebo for 8 weeks and reported a 20 percent reduction in eye wrinkle volume compared with placebo.

A 2019 trial of 72 women used a drink with 2.5 grams of collagen peptides plus acerola, vitamin C, zinc, biotin and vitamin E for 12 weeks and reported better hydration, elasticity, roughness and density than placebo. Because that product bundles several ingredients, it cannot tell you what 2.5 grams of collagen alone does.

The pooled picture: a 2021 meta-analysis of 19 randomized trials (1,125 participants, 95 percent women) found hydrolyzed collagen improved hydration, elasticity and wrinkles versus placebo over roughly 90 days. A 2023 meta-analysis of 26 trials with 1,721 participants confirmed the hydration and elasticity gains while flagging bias in several included trials.

We rate the skin dose range as well-supported, with two caveats: most trials were sponsored or co-authored by peptide manufacturers, and the participants were overwhelmingly women aged 35 to 65, so results in men and younger adults are thinner. For what “improved elasticity” looks like week by week, see our collagen results timeline.

How much collagen for joints? Two products, two doses

Joint supplements come in two forms that are often confused, and they were tested at amounts about 250 times apart. Hydrolyzed collagen (also called collagen peptides, meaning the protein has been enzymatically broken into short fragments that dissolve and absorb easily) was tested at 5 to 10 grams a day. Undenatured type II collagen (intact cartilage collagen, usually from chicken sternum, sold as UC-II) was tested at 40 milligrams a day. They are not a strong and weak version of the same thing.

Hydrolyzed collagen, 5 to 10 grams. In the 2017 Zdzieblik trial, 139 physically active adults with activity-related knee pain took 5 grams of a branded peptide or placebo for 12 weeks; activity-related pain fell more in the collagen group (a 19.5 point drop on a 100 point scale versus 13.9 for placebo), although pain at rest did not differ significantly. In the 2008 Penn State trial, 147 varsity and club athletes were randomized to 10 grams of a liquid collagen hydrolysate or placebo for 24 weeks; the 97 who could be analysed showed less joint pain at rest and during walking, standing and running.

A 2019 meta-analysis of placebo-controlled trials in people with osteoarthritis found lower total WOMAC scores and pain scores with collagen, but the WOMAC pain and function subscores on their own were not significantly different, so the picture is mixed. We rate this as some evidence: positive trials exist, but they are few, most were sponsor-involved, and the effect sizes are modest.

Undenatured type II, 40 milligrams. The idea is not to supply building blocks but to expose the gut’s immune tissue to a small amount of intact cartilage collagen. In a 2016 trial, 191 adults with knee osteoarthritis took 40 mg of UC-II, glucosamine plus chondroitin, or placebo for 180 days; the UC-II group reported lower total WOMAC symptom scores than both other groups. An earlier 2013 trial in 55 healthy adults with exercise-related knee discomfort found better knee extension after 120 days at the same 40 mg dose. Both trials had authors employed by the ingredient’s maker, so we also rate this as some evidence.

If a joint product lists its collagen in milligrams and the number is around 40, it is almost certainly undenatured type II, and comparing it with a 10 gram peptide scoop makes no sense.

Our guide to collagen for joint health goes further into the types and the trials.

How much collagen for bone and nails?

Bone: 5 grams, one 12-month trial. The 2018 Konig trial gave 131 postmenopausal women with age-related low bone density 5 grams of a branded collagen peptide or placebo daily for a year. Bone mineral density at the spine and femoral neck rose in the collagen group relative to controls, and a bone formation marker increased. This is a single trial from one research group with a Collagen Research Institute co-author, so we call the 5 gram bone dose some evidence, not well-supported.

Nails: 2.5 grams, open label. The only dose-specific nail study we could verify is a 2017 open-label trial in 25 people with brittle nails who took 2.5 grams of Verisol for 24 weeks. Nail growth rose about 12 percent and broken nails fell about 42 percent, but there was no placebo group, so expectation and seasonal changes could explain part of the result. We rate it preliminary. For hair, we found no trial that tested one collagen dose against another, so any hair dose you see online is borrowed from skin research.

How much collagen for muscle? 15 grams, always with training

Three placebo-controlled trials gave 15 grams of collagen peptides a day during a 12-week supervised resistance-training program, three sessions a week. In 53 older men with sarcopenia (2015), the collagen group gained about 4.2 kg of fat-free mass versus 2.9 kg on placebo and showed larger strength gains. In 57 recreationally active young men (2019), fat-free mass rose more with collagen, but strength gains were only a trend and muscle fiber size did not differ. In 77 premenopausal women (2019), fat-free mass and hand-grip strength improved more with collagen.

The dose here is inseparable from the training. Nobody has shown that 15 grams of collagen without lifting changes body composition, and two of the three trials came from the same German group using the same branded peptide. We rate the 15 gram muscle dose as some evidence, with training, and we would not call it a reason to take 15 grams for skin.

Can you take too much collagen?

No controlled trial in this review tested more than 15 grams a day, so the honest answer is that doses above that are untested rather than proven safe or unsafe. Within the studied range, the trials reported no serious adverse events; the side effects people describe at these doses are mild and digestive, such as fullness, bloating or loose stools. Our collagen side effects guide covers tolerance, allergies to fish or egg sources, and who should check with a doctor first.

You will see the claim that up to 36 percent of your daily protein can come from collagen. That figure comes from a 2019 calculation using protein-quality scoring on a standard American diet, not from a safety trial, and one of its authors works at the Collagen Research Institute. It does make a useful point: collagen contains no tryptophan and is low in several essential amino acids, so it cannot replace most of your protein, and the same paper notes that the doses shown to work (2.5 to 15 grams) are far below that ceiling anyway.

Anyone with kidney disease, a protein-restricted diet, or a condition affecting amino acid metabolism should ask their doctor before adding any protein supplement. The practical guidance on what not to combine it with is in what to avoid when taking collagen.

What the trials did not test

A dose table can imply more certainty than exists, so here is what the evidence does not cover:

  • Body weight and sex. Every trial above gave a fixed dose to everyone. None adjusted for weight, and most enrolled women aged 35 to 65. Weight-based collagen dosing is an extrapolation, which is why we do not offer one.
  • Doses above 15 grams. Ranges such as “10 to 40 grams” or “20 grams for everyone” appear on brand blogs, but no placebo-controlled trial in our set used them. We rate those ranges as a marketing claim.
  • Splitting the dose. Trials gave one daily dose. Whether two half doses work as well has not been tested, and timing was not part of any protocol; see our note on when to take collagen.
  • Generic products. Most of the positive trials used one of a few branded peptides (Verisol, the Freiburg group’s specific collagen peptides, UC-II, CH-Alpha) and were co-authored or funded by their makers. A generic hydrolyzed bovine collagen is chemically similar, but it has not been through the same trials.
  • Hair and general “anti-aging”. No dose-finding trial exists for hair growth. Claims of a hair dose are borrowed from skin trials.
  • Long-term use. The longest trial here is 12 months (bone). Multi-year safety data at supplement doses do not exist.

Turning a trial dose into scoops, capsules or shots

Once you know which row of the table fits your daily collagen intake goal, the label does the rest:

  1. Find the collagen amount per serving, not per capsule, and convert milligrams to grams by dividing by 1,000.
  2. Powders in our catalogue deliver roughly 9 to 20 grams per serving: Ancient Nutrition Multi Collagen lists 9 g per scoop, Sports Research Collagen Peptides 11 g per scoop and Vital Proteins Collagen Peptides 20 g per two-scoop serving. One 11 gram scoop covers the 2.5 to 10 gram skin range and the 5 to 10 gram joint trials; the 15 gram muscle protocol would need about a scoop and a half.
  3. Tablets and capsules carry far less per piece. The Youtheory label above works out to roughly 1,000 mg per tablet, so its 6 gram serving is six tablets and reaching 10 grams would take ten. This is why our how to take collagen guide favors powder for anyone targeting the joint, bone or muscle doses.
  4. Liquid collagen varies widely; the Double Wood shot above lists 2,500 mg (2.5 g) per serving. Read the number rather than the “high potency” wording.
  5. Undenatured type II products will say 40 mg. That is the tested dose for that product; do not multiply it to chase the gram figures.

The bottom line

The question “how much collagen per day” has no official answer, so the best available one is what the trials used: 2.5 to 10 grams of hydrolyzed collagen for skin, 5 to 10 grams for joints, 5 grams for bone, 2.5 grams for nails and 15 grams for muscle with resistance training, or 40 milligrams if the product is undenatured type II. Convert any milligram figure to grams before comparing it, and treat “10,000 mg” as the ordinary 10 grams it is. Most of these trials ran 8 to 24 weeks with manufacturer involvement, so judge your own results on the same timescale and keep expectations modest.

Frequently asked questions

How much collagen should you take per day?

There is no official daily amount. Clinical trials used 2.5 to 10 grams of hydrolyzed collagen per day for skin outcomes, 5 to 10 grams for joint discomfort, 5 grams for bone density and 15 grams for muscle gains with resistance training. Undenatured type II collagen is a different product and was tested at 40 milligrams.

Can I take 10,000 mg of collagen a day?

10,000 mg is 10 grams, which is inside the range the skin and joint trials used, so it is an ordinary dose rather than a high one. The milligram figure is a labeling choice that makes the number look larger. Controlled trials in our review tested up to 15 grams a day without reporting serious side effects.

Is 2,000 mg of collagen a day enough?

2,000 mg is 2 grams, which sits below the lowest dose used in the placebo-controlled skin trials (2.5 grams) and well below the 5 to 10 grams used for joints and bone. The exception is undenatured type II collagen, a different product that was tested at 40 milligrams. Check which form your label describes.

How many grams of collagen per day for skin?

The placebo-controlled skin trials used 2.5 to 10 grams of hydrolyzed collagen a day for 8 to 12 weeks. In the 2014 trial that compared 2.5 and 5 grams directly, both doses improved skin elasticity against placebo and the higher dose did not add a measurable benefit.

Can you take too much collagen?

No regulator has set an upper limit, and controlled trials in our review did not test above 15 grams a day. Reported side effects at studied doses were mild, mainly fullness, bloating or loose stools. Collagen lacks tryptophan, so it should not replace most of your protein. Anyone with kidney disease or a protein-restricted diet should ask their doctor first.

What are the first signs that collagen is working?

In the trials, the earliest measured changes were in skin: eye wrinkle volume was lower than placebo after 4 weeks and elasticity after 8 weeks in the 2014 trials at 2.5 grams a day. Joint trials measured at 12 to 24 weeks and the bone trial at 12 months. The trials did not record earlier subjective signs.

What should you avoid when taking collagen?

The trials give little guidance here because they tested collagen on its own. The practical points are to avoid judging a milligram label without converting to grams, avoid doses no trial has tested, and avoid treating collagen as your main protein source, since it lacks tryptophan. Our guide on what to avoid when taking collagen covers the rest.

Does the dose change with age, sex or body weight?

The trials did not dose by body weight or sex. They gave every participant the same fixed amount, usually 2.5, 5, 10 or 15 grams, and most participants were women aged 35 to 65. Any site that gives you a weight-based collagen dose is extrapolating beyond what has been tested.

Sources

Every health claim in this article is traceable to a primary source. See our editorial standards.

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
Collagenium

Collagenium

Research-driven editorial team

Editorial & Research Team

Collagenium synthesizes peer-reviewed studies on collagen, skin health, and connective-tissue science into reader-friendly guides. We cite every claim and are not licensed medical professionals. See /methodology/ for our editorial standards.