Skip to content
Types of Collagen Explained: I, II, III and Multi-Collagen
The body has 28 types of collagen; five (I, II, III, V and X) appear on supplement labels. In the body the type matters: type I builds skin, bone and tendon, type II cartilage, type III sits beside type I in skin and vessels. On a peptide label it means less, because a "type" is defined by the triple helix, and hydrolysis takes that helix apart. So "types I and III" means hide, "type I" alone usually bone or fish skin, "type II" cartilage, and a five-type listing a blend of raw materials. No trial has compared types as hydrolysates on any outcome; the skin, joint, bone and training trials almost all used type I peptides (with type III from hide), at 2.5 to 15 g a day.11 The exception is undenatured type II, a 40 mg ingredient that is not a peptide and has its own trials. This page is the type-number explainer for the Wellgard collagen guide.
How many types of collagen are there?
Twenty-eight. Every member of the family has at least one triple-helical domain, and the types differ in which chains they use and how the molecules assemble into fibres, networks or sheets.1 Type IV is the sheet-like basement membrane under the epidermis, type V a minor fibril collagen inside type I fibres, type X a growth-plate collagen; none of the rest is a food ingredient.
What do types I, II and III do in the body?
Type I forms the rope-like fibrils of the dermis, tendon and ligament and the organic framework of bone. Type II is the collagen of joint cartilage and the gel of the eye; skin and bone contain almost none. Type III is type I's companion in skin, vessel walls and gut; in skin its proportion is highest in the youngest tissue and the upper dermis and declines with age, the origin of the "type III for young skin" idea.2 It is never found alone.
Where do the types in supplements come from?
The tissue sets the type. Hide (cattle or pig skin) is type I with type III; bone is type I alone; fish skin and scales are type I, which makes marine against bovine collagen a comparison of two type I ingredients; chicken sternum cartilage is type II; eggshell membrane, found in some blends, carries I, V and X. For hide against bone, see what bovine collagen peptides are.
| Type | Where it is in the body | Supplement raw material | What the trials used |
|---|---|---|---|
| Type I | Skin, bone, tendon, ligament, cornea | Bovine or porcine hide and bone; fish skin and scales | Type I peptides in the skin, knee, bone and training trials, 2.5 to 15 g a day3, 4, 5, 6, 7, 11 |
| Type II | Joint cartilage; the gel of the eye | Chicken cartilage, kept intact as undenatured type II | 40 mg a day in knee trials; no separate record as a hydrolysate8, 9 |
| Type III | With type I in skin, vessels and gut | Hide, together with type I | Arrives with type I in every hide peptide; never trialled alone2 |
| Type IV | Basement membranes | Not a food ingredient | None |
| Types V and X | V inside type I fibres; X in the growth plate | Eggshell membrane | None as a type |
Does the type matter once collagen is hydrolysed?
Not in the way the label implies. Hydrolysis unwinds the helix and cuts the chains into peptides of roughly 2 to 5 kDa, digested into fragments of two and three amino acids; the structure that made the molecule "type I" or "type III" is gone. That is why undenatured type II carries antigenic structures that hydrolysed collagen does not.8 What collagen peptides are explains the hydrolysis step and the kilodalton figure.
So on a peptide label the type is a mark of origin. What we do not know is whether small sequence differences between types matter after hydrolysis; no trial has tested one type's hydrolysate against another's. What the trials show is that hydrolysed collagen works as a class: across 26 randomised trials in 1,721 people, skin hydration and elasticity improved against placebo, and elasticity did not differ by collagen source (p = 0.21).3
Is type II collagen different?
Yes, when undenatured. Undenatured (native) type II collagen is chicken cartilage collagen kept intact, taken as a 40 mg capsule, proposed to work through oral tolerance in the gut rather than by supplying peptides.8 In a 180-day trial of 191 people with knee osteoarthritis, 40 mg a day reduced total WOMAC score more than placebo (p = 0.002) and more than glucosamine with chondroitin (p = 0.04), with safety outcomes that did not differ between groups.9 Powders labelled undenatured type II differ markedly in fibre size and antigenic configuration, so a trial result belongs to the material tested.8 The best type of collagen for joints compares the two joint ingredients.
Do multi-collagen powders do more?
A "multi-collagen" or "collagen complex" listing types I, II, III, V and X is a blend of raw materials, typically bovine, chicken, fish and eggshell membrane. It can mean several lawful sources combined, with the grams of collagen per serving the nutrition panel states; it cannot mean five sets of results, because the types are hydrolysed into peptides together and no trial has tested a multi-type blend against a single-source peptide. The one combination trialled, undenatured type II plus hydrolysed collagen in 68 people with knee osteoarthritis, showed no difference from placebo at 12 weeks, a non-significant result.10 A blend is a legitimate product; the figure to compare across every tub is grams of collagen per serving, and a type count does not change it.
Which type of collagen is best for skin, joints or bones?
The trials decide, not the type number, and in almost every trial the hydrolysate was type I, with type III wherever the source was hide.
- Skin. The 26-trial meta-analysis pooled hydrolysed collagen from several animal sources; hydration gains varied by source, elasticity gains did not.3 In 69 women aged 35 to 55, 2.5 g and 5 g a day for eight weeks both raised elasticity against placebo, and the gain was still measurable four weeks after stopping in the older women.4 See which collagen is best for sagging skin.
- Joints. A 2024 meta-analysis of 35 randomised trials in 3,165 people with osteoarthritis found collagen derivatives reduced pain (SMD -0.35, moderate certainty) and improved function (SMD -0.31, high certainty), with no more adverse events than control.5 The review pooled hydrolysed collagen and undenatured type II as "collagen derivatives".
- Bones. In 131 postmenopausal women, 5 g a day of collagen peptides for 12 months raised spine T-score 0.1 against a fall of 0.03 on placebo (p = 0.030) and femoral neck 0.09 against a fall of 0.01 (p = 0.003).6 What type of collagen is best for bones covers that trial's limits.
- Training. Across 19 trials in 768 adults, collagen peptides taken through eight weeks or more of training improved fat-free mass (SMD 0.48) and maximal strength (SMD 0.19).7
So the best type is the one that was in the trial, at the trial's dose, and for hydrolysed collagen that is nearly always type I: a hide peptide labelled "types I and III" is the trial material, and so is a bone- or fish-derived "type I". Choose by dose per serving, source, allergen status and dietary rules, then allow eight to twelve weeks for skin and joints and a year for bone. Collagen has no authorised health claim in Great Britain; the trials are reported here as evidence, not a promise.
Frequently asked questions
Is it okay to take type 1, 2 and 3 collagen together?
Yes. Types I and III arrive together in any hide peptide because that is the natural make-up of skin.2 Adding type II adds a second raw material: cartilage peptides or undenatured type II at 40 mg. The one trial combining undenatured type II with hydrolysate did not beat placebo at 12 weeks and reported no adverse effects: no evidence of adding up, none of harm.10
What do "collagen 1 2 3" or "types I and III" on a tub mean?
Raw-material statements. "Types I and III" means hide; "collagen 1 2 3" means hide plus a cartilage source, usually chicken. Neither tells you the dose; grams of collagen per serving is the figure that connects a tub to the trials.
Is marine collagen type 1?
Yes. Fish skin and scales are type I, like hide and bone, and skin elasticity in the 26-trial meta-analysis did not differ by source.3 "Marine is type II for joints" is a mix-up: type II comes from cartilage, and fish skin contains none. For plant-based "collagen", see is there a vegan collagen.
Who should not take type 2 collagen, and does it have side effects?
In the 180-day knee trial of 191 people, safety outcomes did not differ between undenatured type II, glucosamine with chondroitin and placebo, and the 12-week combination trial reported no adverse effects.9, 10 It is made from chicken cartilage, so anyone avoiding poultry should check the label, and anyone with a diagnosed joint condition or on regular medicines should mention any supplement to their GP or pharmacist.
Sources
- Ricard-Blum S. The collagen family. Cold Spring Harbor Perspectives in Biology, 2011. PubMed 21421911.
- Ramshaw JA. Distribution of type III collagen in bovine skin of various ages. Connective Tissue Research, 1986. PubMed 2938884.
- Pu SY, et al. Effects of oral collagen for skin anti-aging: a systematic review and meta-analysis. Nutrients, 2023. PubMed 37432180.
- Proksch E, et al. Oral supplementation of specific collagen peptides has beneficial effects on human skin physiology: a double-blind, placebo-controlled study. Skin Pharmacology and Physiology, 2014. PubMed 23949208.
- Liang CW, et al. Efficacy and safety of collagen derivatives for osteoarthritis: a trial sequential meta-analysis. Osteoarthritis and Cartilage, 2024. PubMed 38218227.
- König D, et al. Specific collagen peptides improve bone mineral density and bone markers in postmenopausal women: a randomized controlled study. Nutrients, 2018. PubMed 29337906.
- Bischof K, et al. Impact of collagen peptide supplementation in combination with long-term physical training on strength, musculotendinous remodeling, functional recovery, and body composition in healthy adults: a systematic review with meta-analysis. Sports Medicine, 2024. PubMed 39060741.
- Harris RB, et al. Functional characterization of undenatured type II collagen supplements: are they interchangeable? Journal of Dietary Supplements, 2022. PubMed 34060435.
- Lugo JP, Saiyed ZM, Lane NE. Efficacy and tolerability of an undenatured type II collagen supplement in modulating knee osteoarthritis symptoms: a multicenter randomized, double-blind, placebo-controlled study. Nutrition Journal, 2016. PubMed 26822714.
- Yuenyongviwat V, et al. Efficacy of combined undenatured type II collagen and hydrolysed collagen supplementation in knee osteoarthritis: a randomised controlled trial. Scientific Reports, 2025. PubMed 40897777.
- Paul C, Leser S, Oesser S. Significant amounts of functional collagen peptides can be incorporated in the diet while maintaining indispensable amino acid balance. Nutrients, 2019. PubMed 31096622.
