Do Collagen Supplements Work?

The claim

Collagen rebuilds your joints, repairs your tendons, and while it is at it, fixes your skin, hair and nails.

This one is unusual for this series, because collagen is not a nothing supplement. There is real trial evidence behind it.

The problem is the gap between what the trials measured and what the tub is sold for, particularly if you bought it as a gym supplement.

Where the evidence is genuinely strongest: skin

A systematic literature review searched three databases from inception to April 2023 for randomised controlled trials of oral hydrolysed collagen against placebo or active comparators in adults, assessing skin hydration, elasticity, wrinkles, transepidermal water loss, safety and quality of life. Risk of bias was assessed with Cochrane RoB 2.

Twenty-five randomised controlled trials were included, with samples ranging from 13 to 236 participants. Most were in healthy volunteers, seven in people with chronic conditions.

Collagen supplementation showed significant improvements over placebo in key skin parameters, with skin hydration improving in 10 of 15 studies that measured it.

That is a real body of randomised evidence. Note also the review's own framing: clinical evidence remains inconsistent. Ten of fifteen means five found nothing.

So if you are buying collagen for skin, you are on the firmest ground available here. Not certainty, but genuine randomised trials pointing the same way more often than not.

Tendons: interesting, and narrower than the marketing

A systematic review searched four databases from inception to May 2025 for randomised controlled trials of collagen supplementation in humans, assessing quality with the PEDro scale.

Of 887 unique citations, eight randomised controlled trials met the criteria, totalling 257 participants aged 18 to 52. The sex split is worth stating: 246 men to 11 women.

The premise is that tendons adapt to mechanical loading by increasing cross-sectional area, stiffness and matrix organisation, and collagen has been proposed to enhance that by supplying amino acids for collagen synthesis. The review's own verdict on why this has not translated: inconsistent results across trials have limited its clinical and athletic application.

The heterogeneity was substantial enough in supplementation protocols, training modalities and outcome measures that the authors synthesised results narratively and did not attempt a meta-analysis.

So the honest summary is that collagen is tendon-curious. Something may be happening in connective tissue, in a small and overwhelmingly male evidence base that the reviewers could not even pool. It is not a muscle supplement.

The mistake that actually costs you

Here is the practical error.

Collagen is a protein, and people count it toward their daily protein target. It is a poor protein for building muscle: it lacks tryptophan entirely and is low in the amino acids most associated with muscle protein synthesis.

So swapping a whey shake for a collagen shake means hitting your protein number on paper while feeding your muscle worse. Adding collagen on top of adequate protein is a different decision, and a defensible one if tendons or skin are your goal.

I am labelling the amino acid point as established nutritional composition rather than a finding from the studies cited here, because none of these three tested that substitution directly.

What the protein comparison actually shows

If collagen is a protein, the fair question is how it ranks against other protein sources when combined with training.

A network meta-analysis and meta-regression searched six databases through December 2025 for randomised controlled trials combining any protein supplementation with resistance, aerobic or multicomponent training in untrained older adults, assessing lean mass, handgrip and leg strength, and physical mobility.

A network meta-analysis is the right tool here, because it lets different protein sources and training types be ranked against each other rather than each tested in isolation.

The existence of that analysis is the point I want to make. Protein source is a variable that changes outcomes, which is exactly why treating all protein powders as interchangeable, collagen included, is a mistake.

That analysis was conducted in untrained older adults for sarcopenia outcomes, so it does not transfer directly to a young lifter, and it did not single out collagen.

So is it worth buying

If your goal is skin, the randomised evidence is the best in this article and it is reasonable to try, with the caveat that a third of the measured studies found nothing.

If your goal is tendon health and you already train, there are hints worth a punt, from a small and almost entirely male evidence base.

If your goal is muscle or strength, no. The tendon review found no added strength benefit over training alone, and collagen is a weak protein source for that purpose.

What should not persuade you is the phrase joint support on the packaging, which is doing an enormous amount of undefined work.