Creatine dosage and evidence
Written by Aaron CuhaReviewed Sep 2026
The short answer
The most studied supplement there is, with a solid strength effect, a smaller hypertrophy effect than marketing implies, and a cognition claim a regulator has formally rejected.
- People given it
- 168
- Human studies
- 3
- Found nothing
- 2
Dose
Approved medicine. No human dose established for unlabelled uses
Labeled amounts sit on the prescribing information. This page records what trials measured, not a personal dose.
Sold as a dietary supplement. Not an approved drug for any indication.
5 effects reported, 3 unwanted, 2 wanted, by people using it, not a trial
In plain English
In plain English, from Aaron
This is how I would explain it to a friend. The evidence, with the numbers, is further down the page.
Creatine helps your muscles push harder for a few more seconds. Your body makes it and you eat it in meat. It is the most studied supplement there is. It makes you stronger, and that part is settled. It is sold as a supplement. It is not a drug and does not need to be.
What people take it for
- Getting stronger.
- Bigger muscles.
- A sharper brain and better memory.
- Holding on to muscle as they get older.
What the trials actually showed
Strength is the claim that holds. Hundreds of trials say it makes you stronger and lets you do more work in a session. Muscle size is smaller than the ads say. When you scan the muscle itself instead of the whole body, the effect is small and might be zero. Brain is a no on current evidence. The biggest trial had 123 people and found no benefit, with more side effects on creatine. The European food regulator looked at a company's own file and rejected the brain claim. Hair loss rests on one study of 20 rugby players. It measured a hormone and never looked at hair. A later 12 week trial in 45 men then measured hair and hormones. Thirty eight finished. Nothing moved. Kidneys: a 12 week trial in lifters on a high protein diet found no change in kidney filtration. And you do not need to load. Three grams a day gets you there in about four weeks.
What people report
Reports, not trial results
The good
- Stronger lifts and a couple more reps within a few weeks.
- Cheap, and almost nobody has to stop because of side effects.
The bad
- Water weight in the first week or two. It goes into the muscle, not under the skin, and it is near universal.
- Stomach upset at loading doses. It goes away at three to five grams a day.
- Hair loss worry. It is everywhere on forums and makes people quit, and it rests on one unrepeated study that never measured hair.
Where these come from: Strength training forums and supplement communities for the water and stomach reports. The side effect count in the brain trial is from that trial's own tables.
My bottom line
This is the real thing. It has the strongest evidence of anything in this category. Three to five grams of plain monohydrate every day, whether or not you lift.
An opinion, not a finding. I am a coach, not a doctor.
What it has been measured to do
Measured in people
Goals Creatine has been measured for in a human being, most people first. Counts come from the study lines on this page and deliberately undercount.
- Memory and thinking123 people · 1 study
- Hairno headcount stated · 1 study
What people using it report
Uncontrolled and self-selected, so it cannot show that anything works. It is still the best guide on this page to what people actually do with Creatine, what they expect, and what goes wrong. The full account, including the negative reports, is below.
32 of the 36 indexed goals have no study of any kind behind Creatine
No trial, no animal work, no cell work on this page reports a result for these. That is an absence of evidence rather than evidence of absence, but it does mean anything you have been told about Creatine and one of these did not come from a study cited here.
When this earns a spot
It matters because the strength and lean-mass literature is large and the safety record in healthy people is unusually clean. It is the one floor item that is not marker-driven. It does not matter as a kidney tonic, and it does not wait for a deficiency.
Owner stack tier 1
Bloodwork that belongs beside the bottle
Optional eGFR context, not a start/stop gate for healthy people. Tell the clinician before the draw.
Default retest window: 90 days. If the named marker did not move, stop.
- Creatinine often rises a little. That can look like kidney trouble if nobody writes down that you take creatine.
Reading the number
There is no creatine blood target to chase. If creatinine jumped and you started creatine, say so. Cystatin C is a clinician option when the picture is messy.
Marker glossary: BMP electrolytes.
What people typically order
A BMP is enough context. You do not need a special 'creatine panel'. Buy the monohydrate, not a branded creatine-adjacent stack, if you buy it at all.
Show full evidence
Overview
The most studied supplement there is, with a solid strength effect, a smaller hypertrophy effect than marketing implies, and a cognition claim a regulator has formally rejected.
Creatine is a compound your body makes and you also eat in meat, stored in muscle as phosphocreatine and used to regenerate ATP during short intense efforts. Its effect on strength and work capacity is real and about as well established as anything in this catalogue. Two things are less well established than the marketing suggests. When muscle growth is measured by direct imaging rather than by whole-body composition, the pooled effect is small and its credible interval crosses zero. And the cognition claim, which now drives a large share of sales to older adults, was formally evaluated by the European Food Safety Authority at the request of a creatine manufacturer, and the regulator ruled the claim was not causally established.
Mechanism of action
Loads muscle phosphocreatine, which regenerates ATP from ADP during high intensity work lasting seconds. That raises work capacity per session, and the extra work done over months is the likely route to most of the strength gain rather than any direct anabolic action.
Human evidence
Extensive. Hundreds of randomised trials, which makes creatine unusual in this catalogue: the question is not whether human data exists but what it actually shows once you separate the outcomes.
- Strength and high intensity work capacity: consistently improved across decades of trials. This is the claim that holds.
- Hypertrophy: when measured by direct imaging rather than whole-body composition, the pooled effect is small and its credible interval crosses zero.
- Cognition in healthy rested adults: the largest randomised trial, 123 people, found no benefit, and the European regulator rejected the claim on a manufacturer's own dossier.
- Kidney function in healthy people: a 12 week randomised trial in resistance-trained adults on a high-protein diet found no change in measured filtration, and no trial has shown harm in healthy people.
- Hair: the 2009 hormone study never measured hair. A 2025 randomised trial that did measure hair and androgens in 45 men found no group difference.
- Loading is optional. Three grams a day reaches the same muscle saturation in about four weeks that twenty grams a day reaches in six days.
What this does not tell you: Nearly every favourable synthesis in this area has authors with industry ties, which is disclosed on the papers and worth knowing. The strength effect survives that scrutiny. The hypertrophy and cognition claims are where the gap between the marketing and the measurement is widest.
Reading the research record
Creatine is the rare case on this site where the problem is not a missing trial record. It is that a large, genuinely positive literature has been stretched to cover claims it does not support. The strength evidence is strong enough that creatine would be worth taking on that basis alone. The hypertrophy effect shrinks when you measure the muscle directly instead of inferring it from a whole-body scan, and the cognition claim was tested by a regulator at a manufacturer's own request and failed.
The hair loss story is the mirror image, a claim that circulates as established fact and rested for years on one 20 person study in rugby players that measured a hormone ratio and never looked at anybody's hair. A 2025 randomised trial that did measure hair and androgens found no change. Believing the claim now requires ignoring that null, which is the same credulity as believing the overclaims, pointed the other way.
The evidence, charted
Fig. 1a · evidence scale
168people, across 2 human studies cited here
- 2023 · BMC Medicine12373%
- 2025 · Journal of the International Society of Sports Nutrition4527%
One study holds 73% of these participants, so the total is less independent than its size suggests. Counts are the enrollment each human citation on this page states, which includes observational cohorts where nobody was given the compound. 2 further human citations state no participant count and are not counted here. Studies still recruiting are excluded, and only studies cited on this page are counted.
Fig. 1b · evidence mix
4 of 6 citations here are human work, the rest are not.
- Human · given to people467%
- Review · summarises other work233%
Counted from the citation list on this page, typed as this page types it. It understates any literature larger than the sources we cite, and a review counts once however many studies it covers.
Fig. 2 · evidence over time
Evidence spans 5 distinct years, 2009 to 2025, counted from the citation list on this page.
Fig. 3 · legal status at a glance
US
Not approved
UK
Not approved
AU
Not approved
CA
Not approved
Not approved in any of the four jurisdictions shown. A jurisdiction's classification is a regulatory fact, not a verdict on the science; see Legal status below for the full text and any notes.
Fig. 5 · molecular identity
- Modality
- Supplement
- Molecular weight
- 131.1 Da
- Half-life
- About 3 hours
- Sequence length
- None on file
Short, 1 to 6 hours
Half-life as stated on file: About 3 hours in plasma; muscle stores deplete over roughly 4 to 6 weeks after stopping
No amino acid sequence is on file for Creatine, which is expected: a supplement is not built from residues.
Key studies & citations
- Review2023The effects of creatine supplementation combined with resistance training on regional measures of muscle hypertrophy: a systematic review and meta-analysis
When hypertrophy is measured by direct site-specific imaging rather than whole-body composition, the pooled effect is small with a credible interval crossing zero. The age moderation runs toward younger adults, which is the opposite direction to how creatine is marketed for ageing.
- Review2024Creatine and improvement in cognitive function: evaluation of a health claim pursuant to article 13(5) of regulation (EC) no 1924/2006
The European Food Safety Authority assessed the cognition claim on a dossier submitted by a creatine manufacturer and concluded a cause and effect relationship was not established.
- Human2023The effects of creatine supplementation on cognitive performance: a randomised controlled study
Creatine did nothing for thinking, and it caused more complaints than the dummy did. 123 people took part, about half of them vegetarian. Tests of mental performance came out the same whichever they were given. More of the people on creatine reported side effects than the people on the dummy. Creatine has good evidence behind it for training. This is the trial that tested the brain claim, and the brain claim did not survive it.
123 participants, roughly half vegetarian. The trial reports no cognitive benefit, with more side effects in the creatine arm than placebo.
- Human2009Three weeks of creatine monohydrate supplementation affects dihydrotestosterone to testosterone ratio in college-aged rugby players
The whole creatine causes hair loss story rests on one small study that never looked at hair. 20 college rugby players took creatine for three weeks. Their blood showed a shift in the balance between two male hormones. Nobody counted a single hair. Nobody followed the players afterwards. And no other team has got the same result in the years since. That is the entire evidence base for a claim repeated everywhere.
The single study behind the entire creatine and hair loss claim. It measured a hormone ratio in 20 rugby players over three weeks. It did not measure hair.
- Human2025Does creatine cause hair loss? A 12-week randomized controlled trial
The first trial that actually looked at hair found nothing. 45 resistance trained men took 5 grams of creatine a day, or a dummy, for 12 weeks. 38 finished. Blood androgens did not separate. Hair density and thickness did not separate. The 2009 hormone study is no longer the last word, and the last word is a null.
45 resistance-trained men aged 18 to 40 randomised to 5 g/day creatine monohydrate or maltodextrin for 12 weeks. 38 completed. Androgens and hair measures did not change between groups.
- Human2013Does long-term creatine supplementation impair kidney function in resistance-trained individuals consuming a high-protein diet?
Creatine did not harm kidneys in lifters who already ate a lot of protein. They took a loading dose and then 5 grams a day for 12 weeks, or a dummy. Filtration measured with a gold standard tracer did not change. This is 12 weeks in healthy trainers, not a lifetime in people with kidney disease.
12-week randomised trial in resistance-trained adults eating at least 1.2 g/kg/day protein. Creatine 20 g/day for 5 days then 5 g/day. Glomerular filtration by 51Cr-EDTA clearance did not differ from placebo (group x time p = 0.64). NCT01817673.
What users report
Self-reported experiences, not evidence. Nothing below was measured under controlled conditions, and reports like these cannot separate a real effect from placebo, from the training or diet change that accompanied it, or from what the product actually contained. They are here because knowing what people describe, including what goes wrong, is worth reading alongside the studies.
- Water retention in the first week or two is near universal and is intracellular rather than subcutaneous.
- Gastrointestinal upset is common at high loading doses and generally resolves at a three to five gram maintenance dose.
- Hair loss anxiety is widespread in forums and drives many people to stop. The 2009 hormone study never measured hair. A 2025 trial that did measure hair found no change.
Sources: Strength training forums and supplement communities. Uncontrolled self-report, included for expectations rather than as evidence.
Frequently asked questions
Has Creatine been tested in people?
The most studied supplement there is, with a solid strength effect, a smaller hypertrophy effect than marketing implies, and a cognition claim a regulator has formally rejected. 3 completed trials, 168 people given it. 168 people given it across 3 completed trials. 2 found nothing.
What is the Creatine dosage?
Approved medicine. No human dose established for unlabelled uses. Labeled amounts sit on the prescribing information. This page records what trials measured, not a personal dose. Sold as a dietary supplement. Not an approved drug for any indication.
Is Creatine FDA-approved?
Sold as a dietary supplement. Not an approved drug for any indication.
Does creatine build muscle?
It reliably improves strength and how much work you can do in a session. The direct muscle growth effect is smaller than usually claimed: when growth is measured by imaging the muscle rather than by whole-body composition, the pooled estimate is small and its credible interval includes zero.
Does creatine help cognition?
Not in healthy rested adults on the current evidence. The largest randomised trial found no benefit, and the European Food Safety Authority rejected the claim on a dossier submitted by a creatine manufacturer. Sleep deprived and vegetarian subgroups are where any signal has been reported.
Does creatine cause hair loss?
The claim started with one study of 20 rugby players that measured a hormone ratio and never looked at hair. A 12 week randomised trial in 45 men later measured both hormones and hair. Thirty eight finished the study. Nothing moved. That is a thin basis for a belief this widespread.
Do I need to load?
No. Three to five grams a day reaches full muscle saturation in about four weeks. Loading gets you there in under a week and causes more stomach upset. Both end at the same place.
Is monohydrate worse than the expensive forms?
No comparison has shown a form that beats monohydrate on a clinical outcome. Monohydrate is the form nearly all the research used and it is the cheapest.