Do collagen supplements work? For skin, the fairest summary is maybe, a little: pooled trials report small improvements in hydration and elasticity after a few months, but the studies are mostly small, many were paid for by manufacturers, and the effect weakens or vanishes when only the most rigorous trials are counted. For hair and nails the evidence is thinner still.
That is not the confident yes of the adverts, nor a flat no. Below we lay out what was measured, in how many people, who funded it, and how to decide if a trial is worth your time and money.
At a glance
Verdict for skin
Possible small benefit for moisture and elasticity. Wrinkle results are mixed.Size of the evidence
One pooled review covered 26 trials and about 1,700 people, roughly 65 per trial on average.Funding issue
Many trials were sponsored by collagen makers. A stricter analysis found no clear effect in high quality or independent trials.Typical trial
Hydrolyzed collagen, several grams a day, for about 8 to 12 weeks, mostly in women.Hair and nails
Very few trials. Not enough to support the claims.Who should ask first
Pregnant or nursing women, people with kidney disease or stones, allergy sufferers, people on medicines.
What collagen trials actually measured
Skin studies mostly tracked four things, using instruments rather than photos.
| Outcome | How it was measured | What pooled results suggest |
|---|---|---|
| Hydration | A probe that reads water in the outer skin layer | Small improvement versus placebo in many trials |
| Elasticity | A suction device that tests how fast skin recoils | Small improvement in many trials |
| Wrinkles | Skin surface scans or graded photos | Mixed. Some trials found shallower lines, others none |
| Dermal density | Ultrasound of the deeper layer | Few trials, uncertain |
A statistically significant change on a device is not always one you would notice. When we say small, we mean small.
How big and how long the studies were
- Number of people. One widely cited meta-analysis pooled 26 randomised trials with about 1,721 participants. That averages out at around 65 people per trial, and some were much smaller.
- Who took part. Mostly women, often middle aged, with fewer men and older adults.
- Dose. Usually 2.5 to 10 g of hydrolyzed collagen daily, sometimes in a drink mixed with vitamins or other ingredients, which makes it hard to credit collagen alone.
- Length. Commonly 8 to 12 weeks, a few up to six months. Long term effects and safety beyond that are poorly studied.
- After stopping. Where trials followed people after they quit, gains tended to fade.
Small, short trials are fine for spotting a signal, but they are easily swayed by chance, by who drops out, and by how results are reported.
The industry funding problem
The editors behind the Harvard Nutrition Source point out that much of the collagen research is funded or carried out by companies that sell collagen. That does not make every result wrong, but sponsored research leans toward positive findings across many fields.
A more recent meta-analysis tested this directly. It pooled 23 randomised trials covering about 1,470 people. Taken together, the trials suggested benefits for hydration, elasticity and wrinkles. But when the analysis was limited to trials judged high quality, or to those without industry funding, the benefits were no longer statistically significant. Industry groups disputed how studies were classified, and a later broad review of many more trials still reported modest gains in elasticity and hydration, with smaller effects in the strongest studies.
Do collagen supplements work beyond skin?
| Claim | Strength of evidence | Notes |
|---|---|---|
| Skin moisture and elasticity | Low to moderate | Modest effect, funding concerns |
| Wrinkle depth | Low | Inconsistent results |
| Hair growth or thickness | Very low | Few, small trials |
| Nail strength | Very low | Small studies, mostly without placebo |
| Joint comfort | Low | Often a different product (undenatured type II) |
| Weight loss | Very low | Protein may help fullness, nothing collagen specific |
See collagen for hair and collagen and weight for detail.
Do collagen creams, masks and serums work?
Spread on the face, collagen behaves differently. Each molecule is a giant compared with the gaps in the outer skin, so it never reaches the dermis, where your own supply sits. A cream, serum or mask built on it acts as a moisturiser instead: a thin film settles on top, binds water, and fine lines soften until you next wash. Nice, and real, yet none of it becomes part of your skin's structure.
If you want a topical that supports collagen, retinoids have much stronger evidence. See retinol for anti-aging, and for collagen products on the skin, collagen cream.
Does vegan or marine collagen work?
Marine collagen is real collagen, mostly type I, and it was used in several trials. It does not appear to work better or worse than bovine sources. Products sold as vegan or vegetarian collagen are different. They contain no collagen at all, only raw materials and helpers, for example vitamin C, glycine, proline or botanical extracts. They have very little direct research behind them. Lab grown collagen made by engineered microbes exists but has hardly been tested in people. Read more on vegan collagen and marine collagen.
Should I take collagen? A decision guide
| If this is you | Then |
|---|---|
| You do not wear sunscreen daily | Start there first. It does far more than any supplement. |
| You use sunscreen and a retinoid and want to try more | A three month collagen trial at 5 to 10 g a day is reasonable. |
| You saw no change after three months | Stopping is sensible. More time rarely helps. |
| You mainly want thicker hair | The evidence is weak. See a doctor about the cause of thinning. |
| You eat little protein | Fix your overall diet first. Collagen is an incomplete protein. |
There is no proven age to begin. Track honestly with the method on our collagen results page.
Questions people ask
Does collagen really work for skin?
Possibly, but modestly. Pooled results point to slightly better moisture and recoil among volunteers who kept a daily dose going for a couple of months or longer. Those gains are less clear in the most rigorous, independent studies.
Does collagen powder really work better than pills?
Powder is not more effective gram for gram. It just makes the doses used in trials easier to reach in one serving.
When does collagen start working?
Main readings in most studies came at the 8 or 12 week mark. If you see no difference by three months, you are unlikely to see one later.
Should I take collagen every day?
Trials used daily doses, and benefits faded when people stopped. If you decide to take it, daily use is the pattern that was tested.
At what age should I take collagen?
There is no evidence for a specific starting age. Most trial participants were adults over 30. Younger adults benefit more from sun protection.
Is collagen good for you?
Healthy adults generally tolerate it well as a protein supplement with possible small skin benefits. It is not essential, and it cannot replace a balanced diet or sunscreen.