Creatine and the brain: what the evidence actually shows

Here is where this sits, stated once and plainly so you can decide whether to read the rest.

Creatine's effect on muscle and strength is one of the best-established findings in sports nutrition. Its effect on cognition is a smaller, noisier and much less settled body of work, built on fewer trials with fewer people, and the gap between that literature and the way creatine is currently marketed for the brain is wide.

There is something there. Memory keeps showing up. But if the reason you are considering creatine is that you read it would sharpen your thinking, the strongest version of the evidence does not support that, and this page is going to show you why rather than sell you the version you came for.

The mechanism is sound, and it is the reason the research exists. Your brain is about two per cent of your body weight and uses around twenty per cent of your energy, in bursts, with very little stored locally. Neurons hold creatine as phosphocreatine and use it to rebuild ATP when demand spikes, exactly as a muscle does. The mechanism is the same one that makes creatine work in muscle, in a different address.

A plausible mechanism is not an outcome, though, and this is where most articles stop and this one keeps going.

This is the single most important fact on the page, and it explains almost everything that follows.

Muscle cannot make creatine. It imports every molecule it holds, which is why topping up the supply moves muscle stores so reliably. The brain is different: it manufactures a good deal of its own, sits behind a barrier that limits what crosses, and appears to defend its own levels. A tissue that makes its own supply is much harder to top up from outside.

That single asymmetry explains why the muscle findings are large and consistent while the cognitive findings are small and argued about. It is not that the cognitive researchers have been sloppy. They are working on a tissue that barely moves.

Nearly every claim you have read about creatine and the brain traces back to one trial from 2003. Forty-five young vegetarian adults took 5 grams a day for six weeks, and the researchers measured working memory and abstract reasoning. Both improved substantially. That paper is where "creatine makes you smarter" comes from.

In 2023 a German team set out to repeat it properly, with a preregistered, placebo-controlled, double-blind crossover design and 123 participants, making it the largest study of creatine and cognition ever run. Same dose, same duration, same two tests.

The most useful thing that trial produced is a number in ordinary units rather than a statistic. On the working memory test, people repeat back increasingly long strings of digits in reverse. The creatine groups managed, on average, a fifth of a digit more.

Whether that is worth anything depends entirely on what you expected. It is not a sharpening you would notice on a Tuesday. It is also not nothing, and the authors make the fair point that a small effect from something cheap and safe, spread across a lot of people and a lot of years, is still worth having.

The wider reviews sit in the same territory. A 2023 meta-analysis of eight trials found memory improving against placebo, with the effect concentrated in participants aged 66 to 76 and essentially absent in the young. A 2024 meta-analysis of 16 trials also found memory improving, along with attention time and processing speed, and graded the memory finding as moderate certainty and the rest as low.

Two caveats on the age story, because it is the part most often repeated and least often examined. The older-adult subgroup in that 2023 review rests on two small studies, only one of which found an effect, and the review left out a trial that had found a large effect in young adults. It also drew a published letter to the editor disputing its statistics, which the authors answered. None of that makes the age pattern wrong. It does mean it is not settled, and that anyone presenting it as settled has not read the argument.

This is the part that should change how you read creatine marketing.

Alongside its two main tests, the 2023 trial ran eight further cognitive tests: sustained attention, visual attention, task switching, visuospatial working memory, verbal learning, visuospatial memory, inhibition and verbal fluency. It found no indication of an effect on any of them. The authors note the spread of results was what you would expect if there were no effect at all.

The 2024 meta-analysis reached the same boundary from the other direction. Across 16 trials it found no significant improvement in overall cognitive function, and none in executive function.

Put those together and the picture is specific rather than general. Creatine has some evidence behind memory, particularly the harder memory tasks. It has essentially nothing behind focus, attention, mental sharpness, task switching or the ability to hold your train of thought while someone interrupts you. Those are the things creatine is currently being sold for, and they are the things it has been tested on and not found to do.

The reasoning is appealing. Dietary creatine comes almost entirely from meat and fish, vegetarians eat none of it, so vegetarians should have the most to gain. The 2003 study was run entirely in vegetarians, which reinforced the idea.

Two things undo it.

Vegetarians do not appear to have less brain creatine. They carry measurably lower creatine in blood and muscle, but imaging studies comparing brain creatine between vegetarians and omnivores have found it comparable. The brain defends its own supply, and diet is not the lever there that it is in muscle.

The trial that tested it directly found no advantage. The 2023 study deliberately recruited about half vegetarians and half omnivores to settle this. Vegetarians did not benefit more. If anything the effect in that group was slightly smaller, though not to a degree worth reading into.

Vegetarians and vegans still have a clear reason to supplement creatine, because their muscle stores genuinely do start lower. That argument holds and it is a good one. It is the brain half of it that does not.

If there is one place the cognitive evidence is least ambiguous, it is performance when someone is short of sleep.

One trial gave 20 grams a day for a week before 24 hours without sleep and found less decline in reaction time, balance and mood. A 2024 trial gave a single large dose during about 21 hours of sleep deprivation and measured both improved cognitive performance and changes in the brain's high-energy phosphates, which is as close as this literature gets to watching the mechanism work.

It fits the mechanism, which is the reason to take it seriously. A tissue running short of energy has more to gain from a bigger energy buffer than a well-rested one does. That is also the honest shape of the claim: creatine may help a brain under load rather than improve a brain that is fine.

A straight answer, with the reasoning visible.

If cognition is your only reason, we would not lean on it. The evidence is real but modest, the largest trial did not reach significance, and the tissue barely fills. Buying a supplement on the strength of a fifth of a digit is not a decision we want to talk anyone into.

If you are taking creatine anyway, the cognitive research is a reasonable bonus to hope for. The muscle and strength case is strong, well replicated and not seriously disputed, and it stands on its own. Anything the brain gets is on top of a decision that already made sense.

If you are older, or regularly short of sleep, the odds are better. Both the memory reviews and the sleep-deprivation trials point that way, and both are the places where a bigger energy buffer should matter most. Neither is settled enough to promise.

What would change our view: a large, preregistered trial in older adults, running long enough to matter, measuring brain creatine alongside cognition rather than assuming it. That study has not been done. When it is, we will write about what it found, including if it finds nothing.

It would be reasonable to ask why a brand that has just spent two thousand words tempering the brain research sells the stuff at all. Here is our co-founder Andrew on why we added it.

Both things are true at once. The research has widened well beyond sport, and the specific brain claims currently attached to it have run ahead of what has been shown. We would rather be the page that draws that line than the one that takes the easy version and gets caught by a reader who checks. If you want the part of the evidence that is not in dispute, the muscle, bone and strength research is where to look.