Creatine timing is one of the most argued questions in supplementation and one of the least supported. The interesting part is not the answer — it is that the study everybody cites for the answer did not find it.
This is research journalism. It reports what the sources state, including the amounts used in trials, each figure carrying its source and date — and it never tells anyone what to take.
The paper that started it
In 2013, nineteen recreational male bodybuilders were randomised to take 5 g of creatine monohydrate either immediately before or immediately after training, five days a week for four weeks, with body composition measured by air displacement plethysmography and strength by 1-RM bench press (PMID 23919405, Journal of the International Society of Sports Nutrition).
Here are the mean changes it reported, before group versus after group:
| Measure | Before training | After training |
|---|---|---|
| Body weight (kg) | 0.4 ± 2.2 | 0.8 ± 0.9 |
| Fat-free mass (kg) | 0.9 ± 1.8 | 2.0 ± 1.2 |
| Fat mass (kg) | −0.1 ± 2.0 | −1.2 ± 1.6 |
| Bench press 1-RM (kg) | 6.6 ± 8.2 | 7.6 ± 6.1 |
The after-training column looks better. Now the sentence from the paper's own results section:
While there were trends, no significant interactions were found.
Fat-free mass and bench press improved significantly over time — in both groups, which is the creatine effect, not a timing effect. The comparison between groups, which is the question the study was built to answer, did not reach significance. Look at the standard deviations: ±1.8 kg and ±1.2 kg around means of 0.9 and 2.0, in groups of nine and ten people.
The conclusion that post-workout is superior comes from a separate analysis using magnitude-based inference, and the paper is explicit that this is what it rests on.
What magnitude-based inference is, and why it is a caveat
Magnitude-based inference estimates how likely a true effect is to be beneficial, trivial or harmful, instead of testing against a null hypothesis. It was widely adopted in sports science because it appeared to extract conclusions from the small samples that field can afford.
It was then publicly contested within that literature. By 2018 the argument had reached the field's main journals — one paper in Medicine & Science in Sports & Exercise is titled Bayesian Methods Might Solve the Problems with Magnitude-based Inference (PMID 30431543), which is a statement about the method having problems before it is anything else. The core objection from methodologists is that it calls trivial effects meaningful more often than conventional testing does.
This page does not need to settle that dispute. The relevant point is narrower and is not in dispute: the widely repeated claim that creatine works better after training traces to one four-week study in nineteen people whose conventional analysis found no significant difference, and whose contrary conclusion depends entirely on a method the field was arguing about five years later.
The design that answers it properly
The strongest test of a timing question is to put both timings in the same person, and that is what a Canadian group did (PMID 34610729, Journal of Sports Medicine and Physical Fitness, 2021).
Ten participants trained unilaterally — elbow flexion and knee extension, 3–6 sets at 80% of 1-RM, for 8 weeks. In a counterbalanced, double-blind, repeated-measures design, each person took creatine (0.1 g/kg body mass) immediately before training one side of the body and placebo after, and on alternate days the reverse for the other side.
Every participant is their own control. Diet, sleep, genetics, training history, motivation, day-to-day recovery — all identical across the two conditions, because they are happening in the same body at the same time. Total training volume was confirmed similar between conditions (P = 0.56).
Muscle thickness by ultrasound and 1-RM strength both increased significantly over time, with no differences between the two ingestion strategies.
And a conventional trial in athletes agrees
A third study ran the parallel-group version properly with a placebo arm: 34 resistance-trained male and female collegiate athletes, matched on fat-free mass, randomised to placebo, 5 g of creatine within an hour before training, or 5 g within an hour after, for 8 weeks, all groups also taking 25 g of whey isolate and 25 g of maltodextrin (PMID 36465581, Frontiers in Sports and Active Living, 2022).
All groups gained fat-free mass (+1.34 ± 3.48 kg, p = 0.04), upper-body strength (+2.21 ± 5.69 kg, p = 0.04) and lower-body strength (+7.32 ± 10.01 kg, p < 0.001), and lost fat mass (−2.64 ± 4.16 kg, p = 0.001).
The authors' conclusion: the timing of creatine monohydrate did not exert any additional influence over the measured outcomes.
What the review concluded, and who funded it
The dedicated 2021 review of the question, in Nutrients, describes emerging evidence of greater benefit after exercise but states that methodological limitations currently preclude solid conclusions, and notes that the physiological and mechanistic data supporting a timing effect are lacking (PMID 34445003).
Its conflict-of-interest statement discloses that authors have received research grants, creatine donations for scientific studies, travel support and speaking honoraria from a named creatine manufacturer, and that one serves on that company's scientific advisory board. The disclosure says the company provided no intellectual input. This site reports such disclosures on every creatine page as a matter of routine rather than as an accusation — a reader is entitled to know who funded the literature they are being shown.
Why a timing effect was never mechanistically likely
Creatine is not a signal. It is a substrate that accumulates in a store.
Muscle total creatine takes days of consistent intake to rise, and once elevated it declines over weeks rather than hours after intake stops — which is the same physiology that makes the loading phase a shortcut rather than a requirement. An intervention that works by the standing level of a slowly turning-over pool is a poor candidate for an effect that depends on landing within an hour of a training session.
That is the honest mechanistic reading, and it is why the reviews note that mechanistic support for a timing benefit is missing rather than merely unproven.
A search trap worth recording
A PubMed search for creatine and timing in the title returns nine records. Two of them are not about supplementation at all: one is on the timing of troponin T and creatine kinase-MB elevations after coronary intervention, and one is on how cerebral creatine deficiency affects the timing of oligodendrocyte myelination in brain development.
That is the same trap this site has now hit repeatedly on this compound — creatine kinase is not creatine, and a count is not a finding until the records are classified. Two of nine is a 22% error rate on a search that looks precise.
The honest summary
Three trials were designed to detect a creatine timing effect on muscle size or strength. The one that reported a difference reported it as not significant and reached the opposite conclusion by a contested statistical route. The one that removed every confound by putting both timings in the same person found nothing. The one with a placebo arm and 34 athletes found nothing.
What fills the muscle store is total intake over time. That is what the evidence supports, and the timing argument is considerably larger than the finding underneath it. The same shape recurs one question over: the evidence on taking creatine during a calorie deficit is two small randomised trials that disagree, against a very large amount of confident advice.
Limits of this page
All three trials are small — 19, 10 and 34 participants — and none was powered to detect a difference smaller than a few hundred grams of fat-free mass. Two of the three enrolled mostly or only men. The within-subject design is the strongest available here but it used unilateral upper- and lower-limb exercises, not whole-body training, and its sample was ten. Absence of a detected difference in trials this size is not proof that no small difference exists; it is the reason no one has been able to demonstrate one.
Sources and dates
- Antonio J, Ciccone V. The effects of pre versus post workout supplementation of creatine monohydrate on body composition and strength. J Int Soc Sports Nutr 2013;10:36. PMID 23919405
- Forbes SC, Krentz JR, Candow DG. Timing of creatine supplementation does not influence gains in unilateral muscle hypertrophy or strength from resistance training in young adults: a within-subject design. J Sports Med Phys Fitness 2021;61(9):1219–25. PMID 34610729
- Dinan NE, Hagele AM, Jagim AR, Miller MG, Kerksick CM. Effects of creatine monohydrate timing on resistance training adaptations and body composition after 8 weeks in male and female collegiate athletes. Front Sports Act Living 2022;4:1033842. PMID 36465581
- Ribeiro F et al. Timing of creatine supplementation around exercise: a real concern? Nutrients 2021;13(8):2844. PMID 34445003 — conflict-of-interest statement read in full.
- Bayesian methods might solve the problems with magnitude-based inference. Med Sci Sports Exerc 2018. PMID 30431543
- NCBI PubMed E-utilities, title-restricted search for creatine and timing, run 13 September 2026.