Walk into any gym forum and you’ll find someone certain that creatine belongs in the shaker before you lift, and someone equally certain it belongs in the one after. Both camps are arguing over a question that has been tested directly in exactly five published trials, and the honest reading of those trials is that the clock doesn’t seem to care.
The short version: no study has found that taking creatine before a workout produces better strength gains than taking it after, or the other way around. Two studies reported slightly larger gains in lean mass with creatine taken after training, but both have problems that the researchers who ran them openly acknowledge, and three later studies found nothing. The scientists who study this most closely now say the same thing. Take a consistent daily dose and stop worrying about the timing.
That conclusion comes from a researcher who went looking for a timing effect with a plausible mechanism in hand. Bryan Saunders of the University of São Paulo presented the case at a creatine conference after coauthoring a 2021 review in Nutrients asking whether creatine timing is a real concern. His answer, after walking through the physiology and the data, was that it isn’t. Here is why, and what has been published since.
What the Evidence Actually Shows
The whole literature is small enough to hold in your head. Five randomized trials have compared creatine taken immediately before resistance training against creatine taken immediately after, over periods of four to thirty two weeks, in a total of a little over 120 people. All five measured strength and found no difference between timing groups. All five measured some version of muscle or lean mass, and two reported a small advantage for creatine after training while three found none. Not one of the five measured creatine inside the muscle, which means none of them can say whether timing changed the thing it was supposed to change.
A 2018 mini meta analysis pooled the first three of those trials and found a statistically significant edge for creatine after training on muscle mass, with a standardized mean difference of 0.52 and a confidence interval running from 0.03 to 1.00. That interval barely clears zero, and the two null trials published since would almost certainly pull it back toward the middle. The most recent expert consensus, a 2025 paper from twenty creatine researchers, concludes that the evidence doesn’t validate timing as critically important.
So the rating at the top of this article is “unsupported.” That doesn’t mean creatine doesn’t work. It means the specific claim that when you take it relative to a workout changes your results is not supported by the studies designed to test it.
Why Anyone Thought Timing Would Matter
The timing idea isn’t silly. It grew out of one of the most important creatine studies ever run.
In 1992, Roger Harris and colleagues published the first human study showing that oral creatine raises muscle creatine content. Participants took 20 to 30 g per day in split doses and increased total muscle creatine by as much as 20 percent. Buried in that paper was a second finding.
Some participants cycled with one leg for an hour a day during supplementation and rested the other, and the exercised leg loaded more creatine than the rested one. A 1999 follow up from the Greenhaff group found a 68 percent greater rise in total creatine in a leg that had exercised compared with the rested leg. Saunders describes this as close to settled: muscle that’s working takes up more creatine than muscle that isn’t.
From there the reasoning goes like this. If exercise helps muscle absorb creatine, maybe having creatine in the bloodstream while the muscle is working helps even more. Two mechanisms make that plausible.
The first is exercise hyperemia, which just means more blood flowing to a working muscle, and with it more creatine delivered. The second involves the sodium and potassium pump in the muscle cell membrane. Creatine enters muscle through a transporter that rides on the sodium gradient, and that pump works harder during contraction, so in theory more creatine could be pulled in.
Saunders points out that the arithmetic on the mechanism actually favors creatine before exercise, not after. A 5 g dose peaks in the blood around one to two hours after you drink it and stays elevated for several hours more, according to the pharmacokinetic work summarized in the Nutrients review. A typical lifting session runs 40 to 90 minutes, so creatine taken before training overlaps with peak blood flow, while creatine taken after gets a shorter overlap because muscle blood flow drops within about 30 minutes of stopping.
If the mechanism were the whole story, you’d predict a before advantage. The two studies that found anything found an after advantage. That mismatch is one reason to be skeptical that the mechanism is doing much at the timescale of a single workout.
There’s a second reason, and Saunders spends real time on it. Timing can only matter while muscle creatine is still rising. At 20 g per day, muscle stores saturate in five to seven days, while at 3 to 5 g per day saturation takes around four weeks, according to the classic loading work by Hultman and colleagues.
Once you’re full, there’s nothing for timing to speed up. So the only window where timing could plausibly matter is the first few weeks of a low dose protocol, and even then the question is whether getting there a few days sooner changes anything you’d notice.
The Four Studies Saunders Reviewed
The 2021 review Saunders coauthored, Timing of Creatine Supplementation around Exercise: A Real Concern?, found four trials that directly compared before versus after. His presentation walked through each one, and the details matter more than the headlines.
The first, by Jose Antonio and Ciccone in 2013, is the one most often cited by people who favor creatine after training. Nineteen recreational male bodybuilders trained five days a week for four weeks, taking 5 g of creatine monohydrate either immediately before or immediately after each session, and whenever they liked on rest days.
Almost every outcome improved in both groups. Body weight, fat mass, body fat percentage and bench press one repetition maximum showed no difference between groups. Fat free mass was the exception: the after group gained about 3 percent, the before group about 1.3 percent. In Saunders’ words, this was the first minor piece of evidence that something might be going on.
He also flags what the paper’s authors flagged. There was no placebo group, so nobody knows how much of either gain came from training alone. Muscle creatine wasn’t measured. Participants knew which group they were in. And the analysis used a statistical method called magnitude based inference, which I’ll come back to, because it changes how much weight that 3 percent deserves.
The second study came from Darren Candow’s group at the University of Regina in 2014. Twenty two healthy older adults new to resistance training did a supervised whole body program three days a week for 12 weeks, taking 0.1 g of creatine per kilogram of body weight either before or after each session, with a matching placebo on the other side of the workout. For most people that dose works out to seven or eight grams on training days and nothing on rest days.
The design was double blind. Leg press and chest press strength went up, muscle thickness went up, lean tissue went up, and none of it differed between timing groups. Saunders was blunt about this one: no difference at all.
The third, also from Candow’s group in 2015, is the largest and longest timing trial ever run. Thirty nine healthy older adults trained for 32 weeks, again at 0.1 g/kg on training days, and this time there was a placebo group. Both creatine groups gained more chest press and leg press strength than placebo, with no difference between before and after. On lean tissue mass, the after group gained significantly more than placebo, while the before group did not separate from placebo. That sounds like a win for after, and Saunders counts it as a second tick in that column.
But he also notes the thing that gets lost in summaries: the before and after groups were not statistically different from each other. One beat placebo and the other didn’t, which is not the same as one beating the other.
The fourth study, published in 2021 by Forbes, Krentz and Candow, used a design I find more convincing than the others. Ten recreationally active young adults trained one side of the body with creatine before and placebo after, and the other side with placebo before and creatine after, on alternating days for eight weeks at 0.1 g/kg. Every participant was their own control, which strips out a lot of the noise between people.
Elbow flexor and knee extensor strength rose. Muscle thickness rose. Neither differed by timing. The paper’s title says it plainly: timing didn’t influence gains in unilateral hypertrophy or strength.
Add it up and Saunders’ summary holds. For strength, zero of four studies found a timing effect. For lean mass, two of four found something in favor of after, and both have caveats that keep them from being decisive.
The Statistics Problem Nobody Talks About
The Antonio and Ciccone result is the foundation of the “take it after” advice, so it’s worth understanding how it was analyzed.
The study reported no significant group by time interaction, which in plain terms means the standard statistical test found no difference between timing groups. The claim that after was better came from magnitude based inference, an approach popular in sports science for a while that classifies effects as “possibly” or “likely” beneficial based on where a confidence interval sits. Kristin Sainani, a Stanford biostatistician, published a detailed critique of magnitude based inference in 2018 showing that it inflates false positive rates, sometimes dramatically, in exactly the small samples where it’s most often used. The Nutrients review notes that the absolute differences in the Antonio study were small and the confidence intervals overlapped.
None of this means the after group didn’t gain more fat free mass. It means a 19 person, four week study with no placebo arm, analyzed with a method that’s since been discredited, is not the kind of evidence you build a rule on.
The same caution applies to the 2018 mini meta analysis by Forbes and Candow, which pooled the Antonio study with the two Candow trials and found that creatine after training beat creatine before on muscle mass, standardized mean difference 0.52, p equals 0.04. Three studies, a confidence interval whose lower bound is 0.03, and no strength effect. Candow was an author on the null within subject study that followed, and his group’s more recent reviews don’t lean on that pooled number, which is what honest updating looks like.
What Has Been Published Since the Talk
Saunders’ review and presentation stopped at four studies. A fifth has been published since, and it’s the largest in young athletes.
In 2022, Nicholas Dinan and colleagues at Lindenwood University ran a randomized, double blind, placebo controlled trial in 34 collegiate athletes from football, soccer and volleyball. Participants took 5 g of creatine monohydrate within an hour before training, within an hour after, or placebo at both times, for eight weeks of supervised off season lifting, with a four week training run in beforehand to even out beginners’ gains. Every dose came with 25 g of whey protein and 25 g of carbohydrate. All three groups gained fat free mass, gained upper and lower body strength, and lost fat, and there were no differences between groups on any outcome.
There’s a second finding hiding in that result. Not only did timing not matter, the creatine groups did not outperform placebo. The authors think the daily protein and carbohydrate, plus the absence of a loading phase, may have masked creatine’s usual effect over eight weeks. It’s also a reminder that with 11 or 12 people per group, an eight week trial can’t reliably detect creatine’s modest contribution on top of training, let alone a difference in when it’s taken.
Candow, Forbes and a group of other creatine researchers, including Saunders’ colleagues Bruno Gualano and Hamilton Roschel, published their own review in 2022 under the title Creatine O’Clock, asking whether timing really influences muscle mass and performance. It covers the same ground and reaches the same place: whether timing influences muscle mass and performance is unclear, and the studies that found an after effect have limitations that prevent solid conclusions.
Then, in a paper published online at the end of 2024, the broadest group yet, twenty authors spanning most of the major creatine labs, published Part II of their common questions and misconceptions series in the Journal of the International Society of Sports Nutrition. Question two on their list was whether timing really matters. Their summary: the current body of evidence does not validate timing as critically important for long term resistance training, before and after appear equally effective for lean tissue and performance, and consistent ingestion is likely the most important variable. When the people who ran the after favoring studies sign a paper saying that, the debate is functionally over until new data arrive.
What the Studies Used Versus What People Actually Do
Something that struck me going through these trials is how little they resemble the way most people take creatine.
Three of the five used 0.1 g/kg, only on training days. For a 70 kg person that’s 7 g three times a week, about 21 g weekly. The typical recommendation, and what most people actually do, is 3 to 5 g every day, around 21 to 35 g weekly, often with a 20 g per day loading week at the start. The Antonio study used 5 g on training days and left rest days to participant discretion. Only the Dinan trial used a daily protocol resembling normal practice, and it delivered creatine alongside enough protein and carbohydrate that the creatine effect itself disappeared.
This matters for two reasons. First, the training day only protocols were never measured for how quickly or fully they saturate muscle, so nobody knows how long the “loading window” during which timing could theoretically matter actually lasted in these studies. Second, none of the trials used a loading phase. If you load, you’re saturated within a week, and the before or after debate becomes irrelevant by the second week of your program.
The practical upshot cuts against the marketing. If timing only matters during the ramp, and a loading phase shortens the ramp to under a week, the simplest way to make the question disappear is to load and then take a maintenance dose whenever you’ll remember it. Loading has its own tradeoffs, mainly a bit of extra water weight in the first days, and it’s worth reading the 2017 International Society of Sports Nutrition position stand on the options.
What Remains Unknown
Saunders is careful to say that “not supported” is not the same as “disproven,” and his list of what’s missing is the most useful part of the talk.
The biggest gap is that no timing study has measured creatine in the muscle. Every one of them inferred an effect, or the absence of one, from strength and body composition, which are noisy downstream outcomes affected by sleep, diet, training effort and measurement error. Saunders argues that a timing effect has to be physiologically real before it can be expected to show up in performance, and that step has been skipped.
The second gap is sample size. The five studies range from 10 to 39 participants. If timing produces a small effect, and it would have to be small given what’s been observed, none of these trials had the power to find it. A null result in a study of 22 people is not strong evidence of no effect. It’s weak evidence of no large effect.
The third is population. Two of the five studies were in older adults, who respond somewhat differently to creatine than younger people. One was in recreational bodybuilders, one in mixed recreational adults, one in collegiate athletes. Saunders raises the question of clinical relevance directly: a small lean mass difference might matter for an older adult trying to hold onto muscle and be irrelevant for a 25 year old, and the data can’t currently distinguish those cases.
Finally, all five studies looked only at resistance training. Nobody has tested timing around sprinting, endurance work or team sport, where blood flow patterns and session lengths differ. And the mechanism itself, whether exercise increases creatine uptake or reduces its loss from muscle, has never been pinned down. Until someone runs the infusion and microdialysis studies Saunders calls for, the physiological story is a hypothesis.
Does Timing Matter for the Brain?
This site is about creatine and cognition, so I’d be leaving out something readers care about if I didn’t address it. The answer is short: there are no studies of creatine timing for any cognitive outcome, and the brain question is almost certainly about dose and duration rather than the clock.
The only relevant data point comes from a 2017 study by Marina Solis and Bruno Gualano at the University of São Paulo, which used magnetic resonance spectroscopy to measure phosphocreatine in both muscle and brain in 64 people, including children, omnivorous and vegetarian adults, and older adults, after seven days at 0.3 g/kg per day. Muscle phosphocreatine rose between 10 and 28 percent depending on the group. Brain phosphocreatine did not change significantly in any group, with shifts ranging from a 0.7 percent decrease to a 3.9 percent increase.
That’s a single week at a loading dose, and whether longer or higher dose protocols raise brain creatine is a separate question with its own literature. The point for this article is narrower. If a week at 0.3 g/kg barely moves brain phosphocreatine, the idea that shifting a 5 g dose by two hours around a lifting session would affect the brain has no plausible basis.
Exercise hyperemia is about blood flow to working muscle. The brain isn’t a working muscle, and creatine entry into it is governed by a transporter at the blood brain barrier that operates on a different timescale entirely. Anyone selling a timing rule for cognitive benefit is inventing it.
Safety and What To Actually Do
Creatine monohydrate at 3 to 5 g per day is, by the standards of any supplement, well studied. The 2021 expert review of common creatine questions concludes it is relatively well tolerated at recommended doses and finds little evidence behind the common worries about kidney harm, hair loss or dehydration in healthy adults. In the United States creatine is sold as a dietary supplement, which means the FDA does not approve individual products or their label claims before sale. If you have kidney disease, take medications that affect the kidneys, are pregnant, or have any condition that makes you unsure, that’s a conversation for your doctor rather than a blog.
On timing, Saunders’ practical advice at the end of his talk matched the 2025 consensus paper: take it at the recommended dose, whenever suits you during the day, and you’re likely to get the benefit. My own reading lands in the same place, with one addition. Attach it to something you already do every day, whether that’s morning coffee, a post workout shake, or dinner, because the thing that actually predicts results in these studies is showing up with the dose consistently for weeks. One of the five trials let participants take creatine whenever they liked on rest days, and three gave nothing on rest days at all, and none of it seemed to matter.
If your shaker is already out after training, take it then. If it’s mixed into your pre workout drink, take it before. Neither choice has evidence against it, and neither has evidence for it beyond two small studies with overlapping confidence intervals. What the research does not support is the idea that you’re leaving gains on the table by picking the wrong side of the workout.
What Would Change the Picture
Saunders lays out the study that would settle this, and it’s not exotic. Take a much larger group than any trial so far, put them on a low daily dose without loading, randomize them to creatine before or after training, and measure muscle creatine directly during the first few weeks, the only window where timing could matter. Pair that with familiarized strength tests and direct muscle size imaging, and you’d have an answer.
Until that study exists, the honest position is the one his group published: adapting creatine timing to when you train is not supported by solid evidence and should not be considered a real concern. That’s a conclusion from a lab that has received creatine from a manufacturer for its research and, by Saunders’ account, publishes regardless of what it finds. The people with the most reason to find a timing effect looked hard for one and didn’t.