Collagen Daily Dose: Hydrolysed Collagen Dosage Guide

There is no universally established collagen dose. Human studies use different proprietary preparations, doses, populations and endpoints, so a dose that produced a result in one trial should be described as a studied dose, not automatically as the correct dose for every hydrolysed collagen product.

Why isn't there one standard dose?

Because trials have used different doses for different outcomes and proprietary preparations, those doses are study protocols rather than interchangeable prescriptions. A meta-analysis of 19 skin trials included doses from 372 mg to 10 g/day [1], while a 12-month postmenopausal bone-density trial tested a specific 5 g/day preparation [2]. Use those doses to understand the evidence behind a particular claim, not as proof that one number is the right dose for everyone.

Studied doses, by goal

Skin: positive trials have used roughly 1–10 g/day depending on the peptide preparation. A 2.5 g/day Verisol trial reported an elasticity effect, but that does not prove 2.5 g/day of every collagen powder is sufficient [3]. Funding- and quality-sensitive reviews also make any universal skin dose difficult to justify.

Bone density: the best-known 12-month postmenopausal trial used 5 g/day of a specific peptide preparation [2]. Treat 5 g as the studied formulation dose, not a general osteoporosis treatment dose.

Joint symptoms: hydrolysed-collagen trials commonly use gram-scale amounts such as 5–10 g/day, but reviews pool different preparations and some trials are negative. A 2025 CollaSel PRO RCT used 10 g/day; that dose cannot be transferred automatically to every product.

Tendon and connective tissue: protocols range widely. The classic exercise study used vitamin-C-enriched gelatin before loading and measured collagen-synthesis markers; newer trials have used 5–30 g collagen with different timing and training programmes. A 2024 training meta-analysis confirms substantial protocol heterogeneity [12].

Muscle-related training outcomes: prominent longer trials used 15 g/day alongside resistance training, while a 2024 mechanistic trial used an additional 50 g/day collagen and still did not increase integrated myofibrillar protein synthesis above the protein RDA [11]. Dose alone cannot compensate for collagen's incomplete amino-acid profile.

General use: a manufacturer's scoop size is a practical serving, not a clinically validated universal dose. Match expectations to the preparation and outcome actually studied.

Why isn't more always better?

There is no convincing clinical dose-response curve showing that more collagen reliably produces a larger benefit. Some trials compare doses, but preparation-specific results should not be generalized to a receptor-saturation threshold. Higher intake adds more protein and hydroxyproline without a demonstrated proportional gain in skin, joint, bone or muscle outcomes.

Is there an established upper limit?

No evidence-based tolerable upper intake level has been established specifically for hydrolysed collagen. Clinical studies have used doses above 15 g/day, including 30 g in tendon protocols and 50 g/day in short mechanistic protein studies, so 15 g/day should not be described as the highest studied dose. Higher doses also contribute more incomplete protein and hydroxyproline, which may matter for people with constrained protein intake or recurrent calcium-oxalate stones.

Direct human evidence shows that 30 g/day gelatin increased urinary oxalate versus whey, and an isolated-hydroxyproline study found higher urinary oxalate at 2,000 mg/day but not at lower studied doses [9] [10]. Neither study establishes a universal maximum collagen dose or proves that ordinary servings cause stones.

Split dose or single dose?

Both schedules are used in practice, but a dedicated clinical comparison showing equivalent outcomes for split versus single daily collagen dosing is lacking. Follow the studied protocol if you are trying to mirror a specific trial; otherwise choose a schedule you tolerate and can maintain. Tendon studies use different doses and both pre- and post-exercise timing, so there is no evidence-based rule that the daily dose must be taken as a single serving exactly one hour before training.

With food or on an empty stomach?

No clinical evidence shows that fasting produces better long-term collagen outcomes than taking it with food. Medication instructions still take priority: do not mix collagen into the fasting window required for medicines such as levothyroxine or oral bisphosphonates. Claims that plain collagen specifically interferes with tetracyclines or iron have not been established; added minerals in a combination product are a separate issue.

Does the dose need to scale with body weight?

Most collagen trials use fixed daily doses rather than weight-based dosing. That reflects how the studies were designed, not proof that body size is irrelevant. Trial doses are therefore useful reference points for the tested populations, not precisely validated weight-based recommendations for very small or very large adults.

What we still don't know

Whether specific molecular-weight fractions require different doses to achieve equivalent outcomes has not been established, and cross-product dose equivalence is often assumed rather than demonstrated. There is also no convincing universal dose-response plateau above 10 g/day. Long-term multi-year dosing strategies, including whether breaks or adjustments matter, remain poorly studied.

Related reading: collagen peptide benefits by indication, timing, and side effects and safety.

Bottom line

Use trial doses as reference points, not universal prescriptions. Skin studies span multiple doses and preparations; 5 g/day is the dose of a specific bone peptide trial; joint studies commonly use 5–10 g/day but are formulation-dependent; and exercise studies range from 5 to 30 g or more. There is no validated collagen-specific upper limit and no evidence that 15 g/day is a hard ceiling. For muscle building, meeting complete-protein needs matters more than escalating collagen.