The honest answer: there’s no single “right” dose, and the number on the label matters less than most people assume. Clinical trials have studied doses from 2.5g to 15g per day and found measurable results at every point in that range — but the meta-analyses that looked across all the trials found something the labels don’t tell you: higher doses didn’t produce better outcomes than lower ones. Duration did.
That’s worth understanding before you compare milligram numbers across products, because the comparison most people make — more mg must mean more effective — isn’t supported by the published evidence.
What the labels say versus what the trials used
Collagen product labels state doses in milligrams, often prominently. The clinical trials that actually tested whether collagen supplementation does anything report in grams. The conversion is simple — 1,000mg equals 1g — but it matters, because the gap between what some products deliver and what the trials used is larger than the label makes it look.
Here are the doses behind the major collagen trials cited across the research, alongside what a single stick of the UK-available jelly-stick products we track delivers:
Doses used in published clinical trials:
| Trial | Dose per day | Duration | What it studied |
|---|---|---|---|
| Proksch et al. (VERISOL skin elasticity) | 2.5g (2,500mg) | 8 weeks | Skin elasticity in 69 women |
| Hexsel et al. (VERISOL nails) | 2.5g (2,500mg) | 24 weeks | Nail growth and brittleness in 25 women |
| Nutrients runners’ knee RCT | 5g (5,000mg) | 12 weeks | Exercise-related knee pain in 180 active adults |
| Clark et al. (Penn State athletes) | 10g (10,000mg) | 24 weeks | Joint pain and mobility in 147 athletes |
| Zdzieblik et al. (muscle strength) | 15g (15,000mg) | 12 weeks | Muscle mass and strength in 53 sarcopenic men |
Labelled collagen per stick in the products we track:
| Product | Collagen per stick | Source type |
|---|---|---|
| BB LAB Pomegranate Collagen S | 1,000mg (1g) | Marine (fish) |
| NOW Foods Collagen Jelly Beauty Complex | 2,500mg (2.5g) | Bovine (VERISOL) |
| Cloo9 Collagen Jelly | 5,000mg (5g) | Marine (fish) |
| FOODOLOGY Collagen Centella Jelly | Not stated on listing | As labelled |
| HAMCHOROK Rg+ Pomegranate | Not stated on listing | Marine (fish) |
YIXO is excluded from this table — it’s a vegan collagen-support blend, not collagen itself, so a milligram comparison wouldn’t be meaningful.
A few things jump out. BB LAB’s 1,000mg per stick is below the lowest dose used in any major published trial (2,500mg). NOW Foods’ 2,500mg matches the VERISOL trials exactly — which makes sense, since it actually uses VERISOL as its collagen source. Cloo9’s 5,000mg matches the runners’ knee trial dose, the only product on this list where a single stick delivers the full studied amount for a joint-related outcome. Two products don’t state a dose at all, which is itself worth knowing before you assume they’re comparable.
Why the milligram number isn’t the whole story
This is where the dose comparison gets less straightforward than “bigger number, better product.”
The 2026 meta-analysis by Batool et al. in the Journal of Cosmetic Dermatology — covering 35 randomised controlled trials of oral collagen peptides — tested whether higher doses produced better skin outcomes. They didn’t find a significant linear dose-response relationship. Studies using 2.5g per day produced improvements in hydration and elasticity. Studies using 10g per day produced improvements in hydration and elasticity. The improvements weren’t reliably larger at higher doses.
What did predict better outcomes was duration. Longer supplementation — twelve weeks rather than four, six months rather than two — was consistently associated with greater improvements. The ARU umbrella review of 113 trials reached the same conclusion independently: longer use, not larger servings.
This doesn’t mean all doses are identical. It means the research, when pooled together, doesn’t show a clear “more is better” pattern. And it means the gap between a 1,000mg product and a 5,000mg product may matter less than the gap between taking something consistently for three months and giving up after three weeks.
What “mg” actually measures on a collagen label
The milligram figure on most UK collagen products is the weight of hydrolysed collagen peptides — the processed, broken-down form of collagen that your digestive system can absorb. Our guide to hydrolysed collagen covers the science of how collagen is broken from its native 300,000-dalton chains into peptides small enough to be absorbed, typically 2,000–5,000 daltons.
But “hydrolysed collagen” and “collagen peptides” on a UK label are synonyms — same ingredient, different names. A product labelling “2,500mg hydrolysed collagen” and one labelling “2,500mg collagen peptides” contain the same thing at the same dose. If you see both terms used in the same product description, they’re not additive.
A few things the milligram figure doesn’t tell you:
- Collagen type. Whether it’s Type I (skin, bone, tendon — the most common in supplements), Type II (cartilage — a different use case), or Type III (alongside Type I in skin and blood vessels) isn’t captured in the dose. Our buying guide lists the type for each product we track.
- Source. Marine (fish), bovine, chicken, or plant-based collagen-support — the dose says nothing about where it came from. Source matters primarily for allergens: marine collagen carries a fish-allergy risk.
- Other ingredients. A 20g jelly stick with 1,000mg of collagen has 19g of other ingredients — pomegranate concentrate, sugar for the jelly texture, flavouring, sometimes added vitamins. The dose is the collagen portion, not the stick weight.
- Bioavailability. Two products with identical milligram counts could theoretically differ in how well the peptides survive digestion and reach the bloodstream. No UK product label reports bioavailability data, and comparative absorption studies between specific products don’t exist in the published literature.
Matching a dose to an outcome
If you’re taking collagen with a specific outcome in mind, the published trials give you a reference point — not a guarantee, but the dose at which a result was actually measured:
- Skin elasticity and hydration: 2.5g per day produced statistically significant improvements in the VERISOL trial at 8 weeks, and the Batool meta-analysis confirmed skin hydration improvements from around 4 weeks across multiple trials at various doses.
- Nails: 2.5g per day for 24 weeks improved growth rate and reduced breakage in the Hexsel et al. trial — a longer duration than the skin studies, at the same dose.
- Joint pain in active people: 5g per day for 12 weeks in the Nutrients runners’ knee trial, and 10g per day for 24 weeks in the Clark et al. Penn State study. Both found significant improvement, but they tested different populations and durations, so this isn’t a direct dose comparison.
- Muscle strength in older men: 15g per day for 12 weeks in the Zdzieblik et al. trial — by far the highest dose, in a very specific population (sarcopenic elderly men combining collagen with resistance training).
None of these outcomes has a dose that’s been shown to be the minimum effective amount. The 2.5g skin dose wasn’t compared against 1g to see whether less still works — the trial tested one dose and found it worked. The evidence tells you what was studied, not where the floor is.
The cost question
Since the meta-analyses say more isn’t reliably better, the practical question becomes: what does each milligram actually cost?
Our Collagen Price Index tracks cost per 1,000mg across UK-available products, including formats beyond jelly sticks. The range is wide — from around £0.04 per 1,000mg for a bulk powder like Nutrition Geeks to over £1.70 per 1,000mg for a low-dose jelly stick.
But if dose size doesn’t predict outcomes and duration does, then the most cost-effective strategy isn’t necessarily the cheapest per milligram. It’s whichever product you’ll actually take every day for months. A more expensive stick you enjoy eating daily may deliver better results than a cheaper powder you abandon after two weeks because you can’t stand the taste or can’t be bothered mixing it. The format-as-consistency-tool argument from our timing guide applies here too.
Can you take too much?
If higher doses don’t work better, is there a ceiling where they become harmful? The short answer: no serious adverse events have been reported in clinical trials using doses up to 36g per day. There’s no established upper limit from any regulatory body, and both EFSA and the FDA classify collagen as safe.
The longer answer involves a genuine biochemical mechanism — collagen’s hydroxyproline content can convert to oxalate, a kidney-stone risk factor — which is covered in detail in our safety guide. At standard supplement doses (2–10g per day), this hasn’t been shown to cause actual harm.
The bottom line
The milligram number on a collagen label tells you how much hydrolysed collagen is in a serving. It doesn’t tell you whether that amount has been shown to do anything specific, or whether a higher amount would do it better. When you see the meta-analyses alongside the product labels, the picture is clear: the trials that worked used doses from 2.5g to 15g, and the pooled data says duration of use is a better predictor of results than dose size.
A 1,000mg stick is below any studied dose. A 2,500mg stick matches the skin and nail trials. A 5,000mg stick matches the joint trials. All three products may deliver similar outcomes if you take them consistently for long enough — or none of them may, if you stop after a week. That’s the honest reading of where the evidence stands.
Our evidence roundup covers the broader case for collagen supplements, and the where-to-buy guide compares doses, prices, and formats across the products we track.