Does 10g of creatine help your brain? What the latest research shows

For thirty years, creatine has been about strength and power. Take 3 to 5 grams of creatine monohydrate a day. It raises the phosphocreatine in your muscle. You get a small but steady gain in high-intensity output and lean mass. This part of the science is about as solid as nutrition gets.

What is new is different. A run of studies asks if a much bigger dose helps the brain. Usually that means 10 g a day or more. The short answer: there is a real signal. It is stronger in some groups than others. And the doses that move the brain are not the ones on most tubs.

Why the usual 5 g dose is a muscle dose

Creatine feeds the phosphocreatine and ATP energy buffer. That buffer handles short bursts of hard work. Muscle takes it up easily from the blood. It uses the CreaT1 transporter. To fill muscle stores takes about 3 g a day for a month. Or load 20 g a day for a week. Then keep up 3–5 g a day.

The brain is a harder problem. It has the same energy demand. It uses the same buffer. But the blood-brain barrier and the brain’s creatine transporter make uptake much slower than in muscle. MR spectroscopy studies show this again and again. Doses that fill muscle do not reliably raise total brain creatine. When they do, the rise is small. Often just a few per cent. Reviews of brain uptake point to a rough threshold. To raise brain creatine, you usually need about 20 g a day for 4–8 weeks. Or 4–10 g a day for several months. Three grams a day is enough for the biceps. It is probably not enough for the cortex.

The Alzheimer’s pilot: a real 11% rise in brain creatine

One study made the brain case most clearly. It is the CABA pilot trial (Creatine to Augment Bioenergetics in Alzheimer’s). Matthew Taylor at the University of Kansas led it. It ran in an Alzheimer’s Association journal in 2025. Nineteen people aged 60–90 with Alzheimer’s took 20 g a day of creatine monohydrate for eight weeks.

• MR spectroscopy showed an 11% rise in brain creatine over the eight weeks. That is bigger than most earlier brain-uptake studies managed with smaller or shorter doses.
• People showed clear gains in working memory. They also came close to a real gain in executive function. Both are hit early in Alzheimer’s.
• The design is the catch. It was a single-arm pilot. No placebo group. A small sample. A short window. The team was clear: the result calls for a bigger trial, not a clinical recommendation.

The interesting bit is not the score change. Pilots like this often move scores the right way. It is the brain-creatine reading. It is one of the cleaner proofs that 20 g a day for two months works in the tissue.

Sleep loss: one big dose, a real effect

A 2024 paper in Scientific Reports tested a one-off big dose. It was 0.35 g/kg. That is about 25 g for a 70 kg adult. Healthy people were kept awake for 21 hours. Then they took cognitive tests. The creatine group did clearly better on processing speed and short-term memory than placebo. MRS readings showed changes in brain energy that fit the result.

Here is the likely reason. Creatine does not make a rested brain smarter. It buffers the energy cost of thinking under stress. So the gap shows up most when the brain is taxed: sleep loss, low oxygen, head injury, or ageing disease.

What 10 g a day does in healthy young adults: “not much, yet”

It would be tidy if 10 g a day helped well-rested 25-year-olds. But the best controlled study says it does not. It was a randomised dose-response trial by Forbes and Candow (Brain Sciences, 2023). It split 30 healthy young adults into three groups: 10 g a day, 20 g a day, or placebo. The trial ran six weeks. They took tests of processing speed, memory, and executive function. They also used fNIRS to track prefrontal cortex activity.

• Neither the 10 g nor the 20 g group beat placebo on cognition.
• The authors did not measure brain creatine. So maybe the dose just did not raise it enough in this group.
• A fair read is this. Healthy, well-rested young adults may already sit near a ceiling on these tests. The benefit of extra brain phosphocreatine only shows when the system is taxed.

Other recent work points elsewhere. A 2024 trial in Nutrients tested 10 g a day of creatine monohydrate with 2 g of guanidinoacetic acid, the direct building block of creatine. Over six weeks it reported faster reaction time, better word recognition, and better Stroop scores. Reviews and meta-analyses through 2024–2025 mostly agree. Across the whole field, the cognitive effect of creatine is small. It is shaky in healthy young adults. It is most reliable in older adults, vegetarians (who start with lower stores), and people under metabolic stress.

What the evidence supports, and what it does not

Read the recent papers as a whole. A careful summary looks like this.

• Muscle: 3–5 g a day of creatine monohydrate is well-proven for strength, power, and lean mass. Going to 10 g a day for muscle alone buys you little. It just fills stores faster.
• Brain, healthy young adults: 10 g a day for six weeks does not reliably help cognition in normal life. Higher doses or longer use might. But the controlled evidence is thin.
• Brain, under stress: there is a real signal for protection under sleep loss and similar strain. It shows at single doses near 0.3 g/kg, or steady higher daily intake.
• Brain, older adults and clinical groups: this is the most interesting signal so far. It includes the CABA pilot in Alzheimer’s, plus small trials in mild cognitive impairment and depression at 5–20 g a day. It is enough to back the bigger trials now running. It is not enough to treat anything.
• Safety: long-term studies at 5 g a day, and shorter ones at 20 g a day, show no kidney or liver harm in healthy people. The early scare stories did not hold up. Existing kidney disease is the clear exception. The main downside of higher doses is GI upset. Splitting the dose across the day usually fixes it.

At a higher dose, the certification question matters more

Creatine monohydrate is easy to make well. The molecule is simple. So is the test. The main quality mark is CreapureⓇ from AlzChem. That refers to a German-made raw material with a known impurity profile. But testing has sometimes found bad stuff in this category. That includes small amounts of dihydrotriazines and other by-products from how it is made.

Two practical points. First, dose. Go from 3 g to 10 g a day, and you triple whatever else is in the powder too. Batch testing matters more at higher intake, not less. Second, sport. Creatine is not a banned substance. But powders are a common source of accidental contamination with substances that are.

The third-party programmes that fit creatine powders best are Informed Sport and NSF Certified for Sport. Both screen every certified batch against the WADA-aligned banned-list. Both audit the factory. Informed Sport is the more common mark on UK-sold creatine.

Practical takeaway

Take creatine for the gym? 3–5 g a day is still the right answer. The research has not changed that. Curious about the brain side? The clearest signals come at doses near 10 g a day or higher, kept up for weeks to months. The people most likely to feel it are those under metabolic stress, vegetarians, and older adults. Not well-rested young people whose cognition already sits near its ceiling.

Whatever dose you pick, one thing holds. The higher you go, the more it matters that the powder is the molecule on the label and nothing else. A batch-tested certification is the cheapest way to check before you scoop.

You can browse all creatine products on Certwell — or filter by Informed Sport or NSF Certified for Sport — and read the longer reference on how each certification compares.