Collagen is the main structural protein of skin, tendon, cartilage and bone. Collagen peptides are collagen from cow, pig or fish hide and bone broken into short chains that dissolve in water. They are sold on the idea that eating collagen helps the body make collagen.

That idea is less silly than it sounds and less proven than the packaging suggests. Digestion breaks the peptides into amino acids and short fragments, some of which appear in blood and are proposed to signal cells to make more collagen. The trials, almost all funded by peptide manufacturers, mostly report positive results, and the reviews that pool them mostly rate the evidence low certainty.

Collagen is low in some amino acids and lacks one essential amino acid entirely, so as protein nutrition it is inferior to whey or meat. The proposed benefit is specific rather than nutritional: certain dipeptides, notably those containing hydroxyproline, survive digestion, reach the blood and are hypothesized to stimulate fibroblasts and chondrocytes to produce collagen and other matrix proteins. Cell and animal work supports the signaling idea. Whether enough reaches skin or cartilage in a person to matter is the open question the trials attempt to answer indirectly.

Skin. A 2021 meta-analysis of 19 randomized double-blind controlled trials in 1,125 people, 95% of them women, found hydrolyzed collagen improved skin hydration, elasticity and wrinkles compared with placebo, with hydration and elasticity confirmed in subgroup analysis, over about 90 days of use.

Joints. A 2019 meta-analysis of randomized placebo-controlled trials in osteoarthritis found collagen reduced the total WOMAC symptom score by about eight points and pain on a visual analog scale, with the WOMAC gain driven by stiffness rather than pain or function.

Muscle, tendon and body composition. A 2015 trial in 53 older men with sarcopenia found collagen peptides with twelve weeks of resistance training produced larger gains in fat-free mass and quadriceps strength and larger fat loss than training with placebo. A 2024 meta-analysis of 19 trials in 768 healthy adults found long-term collagen with training improved fat-free mass, tendon morphology, muscle architecture, maximal strength and recovery, with moderate certainty for body composition and low or very low certainty for everything else.

The pattern across all of it: consistently positive, mostly industry-funded, small trials, and reviewers who rate what they pooled as low certainty.

Collagen peptides are a food, and the trials report them as well tolerated, with occasional fullness, bad taste and stomach upset. Marine collagen is unsuitable for people with fish allergy. As animal-derived products their quality depends on sourcing; contaminant testing is the buyer's responsibility, since no regulator does it before sale. There is no reason to expect, and no evidence of, serious harm at supplement doses.

Collagen peptides are sold as foods and dietary supplements, with no regulatory review of skin or joint claims. In the European Union the authority that assesses food health claims has not authorized collagen claims about skin or joints, so products sold there rely on wording that stays below a health claim. Nothing is restricted in sport. It is not a medicine anywhere.