For people with knee osteoarthritis, multiple RCTs and a meta-analysis show that collagen significantly improves pain and function compared with placebo. The evidence is strongest for hydrolysed collagen peptides (10 grams per day) and for undenatured type II collagen (40 mg per day), although the study on the latter type was partly conducted by manufacturer-affiliated researchers. For muscle building or use by healthy people without complaints, the evidence is too limited to base a recommendation on.
Collagen supplementation for joints has been studied more rigorously than many other dietary supplements. A meta-analysis of 11 randomised controlled trials (870 participants) shows that hydrolysed collagen, also known as 'collagen peptides', significantly improves both pain and joint function in knee osteoarthritis compared with placebo. The effects are statistically and clinically meaningful, but the large variation between the individual studies makes it difficult to pin an exact number on the effect size. That is a fair caveat: the average effect is positive, but certainty about its precise magnitude is limited1.
Alongside hydrolysed collagen there is also undenatured type II collagen (sold as UC-II, 40 mg per day). In an RCT with 191 participants, this type outperformed both placebo and the popular combination of glucosamine plus chondroitin after six months on pain, stiffness, and function as measured by the WOMAC score. This is a striking result, but two of the three authors of that study were affiliated with the manufacturer of the supplement, which makes the findings less independent2. UC-II and hydrolysed collagen are not chemically identical and also work through different mechanisms: UC-II dampens joint inflammation via the immune system, while collagen peptides deliver bioactive compounds that can reach cartilage tissue3.
There is also evidence for athletes without established osteoarthritis. A 24-week RCT in collegiate athletes with joint pain showed significant pain reduction at multiple time points with 10 grams per day of collagen, with the strongest effect on knee pain. One caveat: only 97 of the 147 participants could be analysed, which limits reliability4. A separate study in active middle-aged adults without osteoarthritis found that 10 grams per day improved daily functioning and pain after six months, but the pain effect was only significant in people who exercise more than three hours per week5. These results in healthy, active individuals are preliminary and based on small studies.
A systematic review of 15 RCTs confirms that joint pain and function show the most consistent improvement with collagen peptides. For muscle recovery and muscle building, however, the benefit is limited: collagen synthesis in tissue increased at 15 grams per day, but muscle protein synthesis did not improve compared with other protein sources. Anyone taking collagen purely for muscle building would be better served by whey protein or another complete protein source6.
Regarding safety: multiple studies and reviews report no significant side effects and good tolerability with long-term use. However, no safety studies covering periods of multiple years are available, so genuine long-term data are lacking7,2,3. Based on the available evidence, collagen supplementation is considered safe for most users.
Evidence based on one meta-analysis of 11 RCTs (PMID 39212129), one industry-funded RCT on UC-II (PMID 26822714), one small sports RCT (PMID 18416885), one RCT in active adults (PMID 37551682), one systematic review of 15 RCTs (PMID 34491424), and two reviews/safety articles (PMID 26267777, 36986062). The meta-analysis (PMID 39212129) has high heterogeneity (I²=75-88%), and PMID 36986062 is from an author affiliated with a collagen manufacturer.