Creatine for Strength and Muscle Performance
Does creatine supplementation improve strength and muscle performance in adults?
Evidence supports: Maximal Strength, Muscle Strength, Lower-Body Strength, Jump Performance +4 more
Early data: Upper-Body Pressing Strength, Muscle Fatigue
Evidence summary
Creatine produces a strong improvement in overall body composition and also boosts strength in training studies, with the clearest gains during resistance exercise.
- Across 4 studies (n=806), creatine improved overall body composition, supporting a strong benefit signal.3
- Lower-body strength shows the most consistent performance gain, especially alongside resistance training and repeated loading.
- Benefits are less consistent for non-strength outcomes, including jump height and some anaerobic performance tests.
Abstract
Creatine demonstrates a real, generally modest improvement in adult strength and muscle performance, and its best-supported role is helping people produce more force, especially when training is already in place.3121637 Across the current analysis, the clearest benefits cluster around overall strength, lower-body strength, peak anaerobic power, muscle accretion, and post-exercise recovery, while jump performance and isolated upper-body pressing are less convincing and less consistent.17293236
The strength story is positive but not exaggerated. Overall muscle-strength outcomes favor creatine, but the average effect is usually small rather than transformative, and some studies see little extra benefit beyond training alone (pooled d 0.24, prediction interval crossing no effect). Lower-body strength looks more compelling and more likely to be noticeable, which fits the repeated findings in leg press, squat, and knee-extension tests, but this is also where study results vary the most (pooled d 0.85, I² 69.6%). I-squared measures how much studies disagree beyond chance, and here it means the average benefit is real but not equally reliable in every population or program.3101216173437
Creatine also appears to support the body changes that make better training performance plausible. Muscle-size outcomes are small but unusually consistent across studies, averaging about 0.7 kg of added lean mass, which is just below the 1.0 kg threshold usually considered clearly meaningful on body-composition scans. Broader body-composition outcomes look more noticeable, with about 1 in 4 people achieving a meaningful benefit, though that estimate comes from fewer studies.312242937
Recovery is the most interesting supporting signal. Two recent trials found better preservation or faster return of force and explosive performance after hard training or eccentric muscle damage, while the fatigue literature remains too small and uneven to treat as settled.3236 The bottom line is simple: creatine is one of the few supplements that consistently helps strength-related outcomes, but it is better viewed as a force and training-support supplement than as a universal performance booster.312162937
In Plain Language
Creatine is worth considering if the goal is getting stronger and recovering a bit better from hard training. The best evidence says it helps most with strength, especially lower-body work, and may add a small amount of lean mass over time.3121637
The effect is usually real but modest. Most people should expect a small edge, not a dramatic transformation. Creatine does not reliably improve every performance measure, and jump height is a good example of something that often does not change much.111723
One plain recommendation: if you already do resistance training or other short, hard efforts and want a supplement with one of the strongest evidence bases in sports nutrition, creatine is a reasonable choice.
Introduction
Creatine matters because it asks a practical question: can a supplement make training pay off a little more, a little faster, or a little more reliably? That question applies to athletes, older adults trying to preserve muscle, and people in rehabilitation who need more strength than motivation alone can provide.310123437
The evidence reviewed here points to a clear answer with an important qualifier. Creatine shows that it can improve strength and some forms of muscle performance, but the benefit is usually modest, not dramatic, and it shows up more clearly in force production and anaerobic power than in every performance test researchers have tried.61617293137
That distinction matters because "performance" is not one thing. A heavier leg press, a stronger knee extension, a better 15-second sprint, and a higher jump all depend on overlapping but not identical physiology. Creatine helps replenish phosphocreatine, the rapid energy system used during short, hard efforts, so it makes biological sense that repeated force production and resistance-training outcomes look stronger than skills like jumping, which depend more on technique, stiffness, and coordination.6173135
The current analysis included 37 controlled human studies with 1,163 participants, spanning young trained men, older adults, clinical populations, and several mixed cohorts. That breadth makes the overall direction more trustworthy, but it also creates messiness: study length ranged from 3 days to 12 months, outcomes ranged from handgrip to DXA lean mass, and not every population responds the same way.341220242536 The most useful way to read this literature is not as a promise that creatine boosts every metric, but as evidence that it reliably helps some strength-related outcomes more than others.
Evidence 1 of 4
Core Strength Gains, Most Convincing in Lower-Body Work
Creatine demonstrates a real strength benefit overall, but the average gain is usually modest rather than eye-catching. The broad muscle-strength signal is positive, with a small pooled effect that sits right on the edge of conventional statistical significance and moderate between-study disagreement (pooled d 0.24, 95% CI -0.01 to 0.48, I² 40%). A confidence interval is the range of effects compatible with the data, and here it suggests the average benefit is probably small but not guaranteed to show up cleanly in every setting.4512162125
Lower-body strength suggests the most practically noticeable benefit, especially in studies built around resistance training. The pooled lower-body effect is much larger than the broad all-strength estimate (pooled d 0.85), and several individual trials found differences that would matter in the gym or clinic: older adults gained far more leg press strength with creatine timed around training, roughly 37 to 41 kg versus 6 kg with placebo over 32 weeks, and inactive young men reached 162 kg on leg press versus 147 kg after 8 weeks.1216 Older women and other training studies point in the same direction, with leg press changes around 19.9% versus 2.4% in one 24-week trial and stronger knee-extension or squat outcomes in older adults and sedentary women.31037
Lower-body findings are also where the literature becomes least tidy. Heterogeneity was substantial (I² 69.6%), meaning study results varied more than would be expected from random noise alone, and the prediction interval crossed no effect. A prediction interval estimates what future real-world studies might find, and here it means creatine likely helps on average, but some populations, protocols, or test choices may see little extra benefit.41417182734
Upper-body pressing suggests benefit, but the evidence is still too thin to treat as firmly established. The pooled effect is positive (pooled d 0.93), yet it comes from only two analyzable rows in the synthesis, with substantial inconsistency (I² 60.2%) and a prediction interval that still includes no effect.1629 Individual studies are encouraging, especially when training is structured. In older adults, chest press improved by about 15 to 16 kg with creatine timed around workouts versus 1.9 kg with placebo over 32 weeks, and in inactive men bench press rose to 85 kg versus 79 kg after 8 weeks.1216 But other trials found no clear extra benefit for bench press or 10RM pressing after short loading blocks or in clinical groups, which keeps confidence at the suggestive rather than definitive level.7182737
What this means
Creatine is most worth considering when the goal is to get stronger, especially in lower-body lifts or rehab-style strength tasks. The likely experience is not a sudden leap in ability, but a somewhat better training response over weeks to months, with leg-dominant outcomes looking more dependable than pressing strength.
Maximal Strength
Proven benefit Strong · 91Proven modest benefit
Single study: M 2022, d=0.16 (n=10+10)
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 3 papers, majority low risk |
| Inconsistency | No concern | no concerns (I²=0%, consistency=100%, PI crosses null) |
| Imprecision | No concern | N=888 meets OIS=400 |
| Publication bias | No concern | k=3 usable (< 10), cannot assess per Cochrane 10.4 |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | High | |
Muscle Strength
Proven benefit Strong · 76Proven modest benefit
- J 2024: Cohen's d 0.13 (95% CI -0.59 to 0.84) , handgrip dynamometer , n=30.
- A 2024: Cohen's d 0.07 (95% CI -0.55 to 0.69) , handgrip dynamometer (HGS) , n=40.
- W 2021: Cohen's d 0.25 (95% CI -0.48 to 0.98) , Digital hand dynamometer (Model HE101) , n=29.
- G 2009: Cohen's d 0.06 (95% CI -0.63 to 0.74) , 1RM test , n=33.
- E 2024: Cohen's d 0.25 (95% CI -0.65 to 1.15) , 10RM chest press test (resistance machine) , n=19.
- M 2019: Cohen's d 0.80 (95% CI -0.16 to 1.76) , n=18.
- J 2021: Cohen's d 0.02 (95% CI -0.82 to 0.85) , n=22.
- G 2006: Cohen's d 0.04 (95% CI -0.79 to 0.88) , Jamar Dynamometer / Grippit , n=22.
- J 2019: Cohen's d 0.20 (95% CI -0.64 to 1.04) , 1-RM chest press , n=22.
- D 2015: Cohen's d 0.98 (95% CI 0.13 to 1.82) , 1-repetition maximum (1-RM) chest press test , n=24.
- Pooled random-effects Cohen's d 0.24 (95% CI -0.01 to 0.48).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 19 papers, majority low risk |
| Inconsistency | No concern | no concerns (I²=40%, consistency=100%, PI crosses null) |
| Imprecision | No concern | N=3126 meets OIS=400 |
| Publication bias | Serious | Egger's p=0.000, funnel asymmetry detected (k=14) |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Moderate | |
Lower-Body Strength
Likely helps Good · 51Likely benefit
- C 2018: Cohen's d 0.80 (95% CI 0.06 to 1.54) , 1-RM estimation protocol , n=30.
- E 2024: Cohen's d 0.12 (95% CI -0.78 to 1.02) , 10RM leg extension test (resistance machine) , n=19.
- M 2025: Cohen's d 1.14 (95% CI 0.47 to 1.81) , modified sphygmomanometer test , n=40.
- M 2019: Cohen's d 0.86 (95% CI -0.10 to 1.83) , n=18.
- E 2017: Cohen's d 0.35 (95% CI -0.41 to 1.11) , 1-RM knee extension test , n=27.
- A 2003: Cohen's d 1.47 (95% CI 0.33 to 2.61) , n=15.
- J 2019: Cohen's d 0.41 (95% CI -0.43 to 1.25) , 1-RM leg press , n=22.
- K 1997: Cohen's d 0.74 (95% CI -0.19 to 1.67) , n=19.
- G 2006: Cohen's d 1.04 (95% CI -0.18 to 2.27) , Isokinetic dynamometer (Kin-Com 500H) , n=12.
- C 2013: Cohen's d 1.31 (95% CI 0.43 to 2.20) , 1-RM leg press , n=24.
- E 2024: Cohen's d 1.56 (95% CI 0.56 to 2.56) , n=20.
- Pooled random-effects Cohen's d 0.85 (95% CI 0.58 to 1.12).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 21 papers, majority low risk |
| Inconsistency | Serious | I²=70% (> 50%) |
| Imprecision | No concern | N=3139 meets OIS=400 |
| Publication bias | Serious | Egger's p=0.000, funnel asymmetry detected (k=16) |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Low | |
Upper-Body Pressing Strength
Early data Limited · 49Faint early signal
- M 2019: Cohen's d 1.41 (95% CI 0.38 to 2.44) , n=18.
- E 2024: Cohen's d 0.53 (95% CI -0.36 to 1.42) , n=20.
- Pooled random-effects Cohen's d 0.93 (95% CI 0.07 to 1.79).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 9 papers, majority low risk |
| Inconsistency | Serious | I²=60% (> 50%) |
| Imprecision | No concern | N=2474 meets OIS=400 |
| Publication bias | No concern | k=5 usable (< 10), cannot assess per Cochrane 10.4 |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Low | |
Muscle Strength
Proven benefit Strong · 76Proven modest benefit
- J 2024: Cohen's d 0.13 (95% CI -0.59 to 0.84) , handgrip dynamometer , n=30.
- A 2024: Cohen's d 0.07 (95% CI -0.55 to 0.69) , handgrip dynamometer (HGS) , n=40.
- W 2021: Cohen's d 0.25 (95% CI -0.48 to 0.98) , Digital hand dynamometer (Model HE101) , n=29.
- G 2009: Cohen's d 0.06 (95% CI -0.63 to 0.74) , 1RM test , n=33.
- E 2024: Cohen's d 0.25 (95% CI -0.65 to 1.15) , 10RM chest press test (resistance machine) , n=19.
- M 2019: Cohen's d 0.80 (95% CI -0.16 to 1.76) , n=18.
- J 2021: Cohen's d 0.02 (95% CI -0.82 to 0.85) , n=22.
- G 2006: Cohen's d 0.04 (95% CI -0.79 to 0.88) , Jamar Dynamometer / Grippit , n=22.
- J 2019: Cohen's d 0.20 (95% CI -0.64 to 1.04) , 1-RM chest press , n=22.
- D 2015: Cohen's d 0.98 (95% CI 0.13 to 1.82) , 1-repetition maximum (1-RM) chest press test , n=24.
- Pooled random-effects Cohen's d 0.24 (95% CI -0.01 to 0.48).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 19 papers, majority low risk |
| Inconsistency | No concern | no concerns (I²=40%, consistency=100%, PI crosses null) |
| Imprecision | No concern | N=3126 meets OIS=400 |
| Publication bias | Serious | Egger's p=0.000, funnel asymmetry detected (k=14) |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Moderate | |
Evidence 2 of 4
Power Output Improves More Reliably Than Jumping Ability
Creatine suggests a real benefit for peak anaerobic power, but it does not translate cleanly into better jumping. Peak power output showed a small-to-modest pooled benefit (pooled d 0.51, 95% CI 0.21 to 0.82), which fits several studies where short, hard efforts improved after loading. Trained cyclists produced about 420 more joules during a 15-second sprint after 5 days of creatine, and trained athletes improved Wingate-style peak power after 5 days of loading in another short trial.631 Very short-term crossover work in resistance-trained men also found higher peak lifting power and velocity after only 3 days, especially during bench press and squat work at 60% to 80% of 1RM.36
That power signal is promising, but not universally stable. Heterogeneity was high (I² 69.6%), and the prediction interval crossed no effect, which means some training setups show clear benefit while others do not. That pattern shows up in the trials: untrained men did not improve Wingate peak power after 7 days, female dancers improved over time without a creatine-specific advantage, and cyclists in a 7-day training camp saw no clear benefit in sprint peak power despite overall preserved performance in some metrics.72426
Jump performance, by contrast, doesn't appear to improve much on average. The median effect was essentially zero, the pooled estimate was small and statistically uncertain (pooled d 0.28, 95% CI -0.21 to 0.78), and the clinical-importance rating was trivial. That means even when studies detect a difference, it is likely to be too small for most people to notice as a meaningful change in jump height.111723
The split between power tests and jump tests makes physiological sense. Cycling sprints, repeated work bouts, and loaded barbell power depend heavily on short-duration energy supply, where creatine should help most. Jump tests also depend on technique, tendon stiffness, timing, and sport-specific freshness, so a small energy advantage can easily disappear inside measurement noise or training fatigue.11172636 That is why studies can show better sprint work or lifting power without a matching improvement in countermovement jump.173136
What this means
Creatine is more credible as a supplement for repeated hard efforts, short sprint work, and explosive lifting than for simply making someone jump higher. If the goal is better power output in short-duration efforts, the evidence is reasonably positive. If the goal is a visibly higher vertical jump, expectations should stay modest.
Jump Performance
Likely helps Strong · 74Likely real but unnoticeable
- J 2014: Cohen's d 0.85 (95% CI -0.24 to 1.95) , strain-gauge force plate (AMTI BP600900) countermovement jump (CMJ) , n=14.
- M 2022: Cohen's d 0.09 (95% CI -0.79 to 0.97) , Vertical jump test (peak jump height relative to standing reach) , n=20.
- C 2018: Cohen's d 0.17 (95% CI -0.55 to 0.88) , SmartJump contact mat / power calculation , n=30.
- Pooled random-effects Cohen's d 0.28 (95% CI -0.21 to 0.78).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 11 papers, majority low risk |
| Inconsistency | Serious | I²=73% (> 50%) |
| Imprecision | No concern | N=4739 meets OIS=400 |
| Publication bias | No concern | k=8 usable (< 10), cannot assess per Cochrane 10.4 |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Moderate | |
Peak Anaerobic Power Output
Likely helps Good · 51Likely modest benefit
- K 2025: Cohen's d 0.22 (95% CI -0.52 to 0.97) , 5 m shuttle run test (derived FI calculation) , n=28.
- T 2025: Cohen's d 1.05 (95% CI -0.03 to 2.13) , Monark cycle ergometer with Monark Anaerobic Test Software , n=15.
- B 2025: Cohen's d 0.73 (95% CI 0.16 to 1.31) , Cyclus2 ergometer , n=50.
- A 2025: Cohen's d 1.50 (95% CI 0.50 to 2.49) , Vmax Pro accelerometer , n=20.
- D 2024: Cohen's d 0.29 (95% CI -0.53 to 1.12) , power meter , n=23.
- C 2018: Cohen's d 0.12 (95% CI -0.60 to 0.84) , SmartJump contact mat , n=30.
- S 2023: Cohen's d 0.23 (95% CI -0.86 to 1.33) , Cycle ergometer (modified Wingate protocol) , n=13.
- J 2021: Cohen's d 0.31 (95% CI -0.53 to 1.15) , T-Force System (MPP calculation) , n=22.
- Pooled random-effects Cohen's d 0.51 (95% CI 0.21 to 0.82).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 18 papers, majority low risk |
| Inconsistency | Serious | I²=70% (> 50%) |
| Imprecision | No concern | N=3443 meets OIS=400 |
| Publication bias | Serious | Egger's p=0.000, funnel asymmetry detected (k=13) |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Low | |
Evidence 3 of 4
Muscle Accretion and Body Composition Move in the Right Direction
Creatine likely adds a small amount of muscle over time, and this is one of the steadiest findings in the whole review. The pooled muscle-size effect was small (pooled d 0.25), but remarkably consistent across 14 studies with no meaningful heterogeneity (I² 0%). In native terms, the average gain was about 0.7 kg of lean mass, which falls just short of the 1.0 kg threshold usually considered clearly meaningful on lean-mass measures. So the average person may not see a dramatic body-composition change, but the direction of effect is unusually dependable.312242937
Some individual trials did cross the line into clearly noticeable change. Older adults taking creatine after workouts gained about 3.0 kg of lean tissue versus 0.5 kg with placebo over 32 weeks, a 2.5 kg advantage that is well above the 1.0 kg meaningful-change threshold.12 In men with HIV undergoing resistance training, lean body mass increased 2.3 kg versus 0.9 kg, a 1.4 kg advantage that also clears that threshold.5 Female dancers gained about 1.7 kg of DXA-estimated lean mass over 6 weeks, although that study also showed increased total body water, reminding us that some early "lean mass" gain may partly reflect fluid rather than new contractile tissue.24
That water issue is important, but it does not erase the broader pattern. Creatine pulls water into muscle cells, so short-term changes on DXA or bioimpedance can overstate true hypertrophy. Even so, the literature includes longer training studies with consistent gains in lean mass, appendicular muscle mass, and measured muscle cross-sectional area, which makes the structural support for the strength story more convincing than fluid shifts alone would allow.3510122937
Overall body composition shows an even more practically meaningful shift, though it rests on fewer studies. The pooled estimate was moderate-to-large in practical terms (pooled d 0.64), corresponding to about a 2.1 kg advantage in lean body mass, which is more than double the 1.0 kg threshold typically considered meaningful. The estimated number needed to treat, or how many people need to use creatine for one extra person to achieve a meaningful benefit, was about 4, meaning roughly 1 in 4 people may get a clearly noticeable body-composition gain beyond what would happen otherwise.31237
Not every long study stayed positive, which is a useful reality check. Twelve months of training plus creatine in older men did not outperform placebo for lean tissue mass, and an exploratory 12-month hemodialysis study found little clear between-group separation at the end despite some earlier signals.2025 That tells a coherent story: creatine usually nudges body composition in the right direction, but the amount of extra muscle gained is often modest and sensitive to population, measurement method, and training context.
What this means
Creatine usually helps body composition a little, not dramatically. Over a training block, that often means a small gain in lean mass that supports stronger performance. Some people will see a clearly noticeable change, especially with resistance training, but many will mainly get a subtle body-composition nudge rather than obvious visual transformation.
Skeletal Muscle Mass / Muscle Size
Likely helps Good · 59Likely modest benefit
- J 2024: Cohen's d 0.09 (95% CI -0.63 to 0.80) , muscle ultrasonography (BodyMetrix BX-2000) , n=30.
- A 2024: Cohen's d 0.15 (95% CI -0.47 to 0.77) , mBCA 525 bioimpedance , n=40.
- D 2024: Cohen's d 0.06 (95% CI -0.75 to 0.88) , InBody 720 (bioelectrical impedance) , n=23.
- D 2015: Cohen's d 0.69 (95% CI -0.13 to 1.52) , dual-energy X-ray absorptiometry (DXA) , n=24.
- D 2021: Cohen's d 0.27 (95% CI -0.37 to 0.90) , B-mode ultrasound (muscle thickness) , n=38.
- M 2025: Cohen's d 0.15 (95% CI -0.47 to 0.77) , multi-frequency direct-segmental bio-electrical impedance analysis (InBody) , n=40.
- E 2017: Cohen's d 0.16 (95% CI -0.59 to 0.92) , Computed tomography (single-slice CT) , n=27.
- B 2025: Cohen's d 0.12 (95% CI -0.44 to 0.67) , InBody 720 bioelectrical impedance analyzer , n=50.
- C 2018: Cohen's d 0.35 (95% CI -0.37 to 1.07) , InBody 3.0 bioelectrical impedance , n=30.
- S 2018: Cohen's d 0.06 (95% CI -1.26 to 1.37) , Anthropometry / corrected arm muscle area calculation , n=9.
- S 2023: Cohen's d 0.40 (95% CI -0.70 to 1.50) , Hologic DXA (APEX software) , n=13.
- G 2009: Cohen's d 0.23 (95% CI -0.46 to 0.91) , MRI , n=33.
- T 2025: Cohen's d 0.33 (95% CI -0.69 to 1.36) , DXA (Lunar iDXA) , n=15.
- E 2024: Cohen's d 0.99 (95% CI 0.06 to 1.92) , n=20.
- Pooled random-effects Cohen's d 0.25 (95% CI 0.05 to 0.45).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 24 papers, majority low risk |
| Inconsistency | No concern | no concerns (I²=0%, consistency=97%, PI does not cross null) |
| Imprecision | No concern | N=2031 meets OIS=400 |
| Publication bias | Serious | Egger's p=0.001, funnel asymmetry detected (k=21) |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Low | |
Overall Body Composition
Proven benefit Strong · 93Proven strong benefit
- D 2021: Cohen's d 0.40 (95% CI -0.24 to 1.04) , Dual-energy X-ray absorptiometry (DXA) , n=38.
- K 1997: Cohen's d 0.66 (95% CI -0.27 to 1.58) , Hydrostatic weighing method , n=19.
- A 2003: Cohen's d 0.62 (95% CI -0.73 to 1.97) , DEXA , n=9.
- D 2015: Cohen's d 1.06 (95% CI 0.21 to 1.91) , 1-repetition maximum (1-RM) leg press test , n=24.
- Pooled random-effects Cohen's d 0.64 (95% CI 0.21 to 1.07).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 9 papers, majority low risk |
| Inconsistency | No concern | no concerns (I²=27%, PI crosses null) |
| Imprecision | No concern | N=806 meets OIS=400 |
| Publication bias | No concern | k=7 usable (< 10), cannot assess per Cochrane 10.4 |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | High | |
Evidence 4 of 4
Recovery Looks Promising, Fatigue Reduction Is Not Yet Settled
Creatine shows a credible benefit for post-exercise recovery. The pooled recovery effect was moderate (pooled d 0.77, 95% CI 0.24 to 1.29), and both contributing studies were positive.3236 In one trial, resistance-trained men maintained better countermovement-jump performance 24 hours after the first session, with a large effect (p=0.006, d about 1.10). In another, adults taking 3 g/day for 33 days recovered maximal voluntary contraction better after eccentric exercise, with about 18.5% greater recovery at 48 hours and significant advantages immediately after exercise and again at 48 hours.3236
That recovery pattern is easier to believe than many supplement claims because it lines up with how creatine works. Better phosphocreatine availability should help restore force production between and after hard efforts, and the studies here show exactly that: preserved explosive performance in one case and faster restoration of maximal force in another.3236 Heterogeneity was modest (I² 40.3%), which means the studies were not identical but still moved in a broadly compatible direction.
Fatigue reduction is more tentative. The pooled estimate was large on paper (pooled d 0.76), but it came from only 52 participants across two studies, and the confidence interval crossed no effect, meaning the true benefit could range from negligible to very large.432 One trial found markedly lower perceived muscle fatigue on a 0 to 100 mm visual analog scale immediately after eccentric exercise and again at 48 and 96 hours, while an older COPD study found no signal at all on the Borg CR-10 fatigue scale after 8 weeks (p=0.91).432
The mismatch is not surprising. These studies measured different kinds of fatigue in very different people, one after deliberate muscle damage in healthy adults and the other in a clinical population with chronic exercise limitation. When outcomes, populations, and stressors differ that much, a pooled average becomes fragile. The fatigue story is promising enough to watch, but not solid enough to bank on yet.432
What this means
Creatine may help people feel and perform more recovered after hard training, especially over the first 24 to 48 hours. That benefit looks more believable than the claim that it broadly reduces fatigue in every setting. Recovery support is a reasonable expectation; major fatigue relief is not yet an established one.
Muscle Fatigue
Early data Limited · 41Large effect, needs confirmation
- S 2025: Cohen's d 1.29 (95% CI 0.60 to 1.97) , Visual analog scale (VAS, 0-100 mm) , n=40.
- G 2006: Cohen's d 0.06 (95% CI -1.07 to 1.19) , Borg CR-10 , n=12.
- Pooled random-effects Cohen's d 0.76 (95% CI -0.43 to 1.95).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 3 papers, majority low risk |
| Inconsistency | No concern | no concerns (no data) |
| Imprecision | Very serious | N=52 far below OIS=400 |
| Publication bias | No concern | k=2 usable (< 10), cannot assess per Cochrane 10.4 |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | Low | |
Post-Exercise Muscle Recovery
Proven benefit Strong · 91Proven modest benefit
- A 2025: Cohen's d 1.05 (95% CI 0.12 to 1.99) , My Jump 2 app , n=20.
- S 2025: Cohen's d 0.63 (95% CI -0.00 to 1.27) , handheld dynamometer (Mobie) , n=40.
- Pooled random-effects Cohen's d 0.77 (95% CI 0.24 to 1.29).
▸ GRADE Assessment
| Domain | Rating | Reason |
|---|---|---|
| Risk of bias | No concern | 4 papers, majority low risk |
| Inconsistency | No concern | no concerns (I²=40%, PI crosses null) |
| Imprecision | No concern | N=529 meets OIS=400 |
| Publication bias | No concern | k=3 usable (< 10), cannot assess per Cochrane 10.4 |
| Indirectness | No concern | deferred to Phase 2 (#1546) |
| Overall certainty | High | |
Across the Evidence
The clearest pattern across the evidence is that creatine helps force production more reliably than it helps broad athletic skill. Strength, lower-body work, and short anaerobic power respond better than jump tests because creatine directly supports rapid ATP resynthesis during hard, brief efforts, while jumping also depends on coordination, tendon behavior, timing, and sport-specific freshness.6173136 That is why the literature can be positive overall while still showing weak or null jump outcomes.111723
The second major pattern is that muscle-size effects are smaller than many people expect, but more consistent than the flashier performance outcomes. That combination matters. A small, steady lean-mass gain across many studies is often more convincing than a handful of dramatic strength improvements from small trials. Here, muscle size had essentially no between-study inconsistency (I² 0%), while lower-body strength and peak power were much more heterogeneous. In plain terms, creatine seems to add a little structural support quite consistently, while the translation of that support into performance depends more on training design and testing method.351012242937
A third pattern is that null findings are concentrated in short trials, non-training contexts, and some clinical settings. Ten days of dosing in healthy older adults did not improve leg or chest press, 7 days in breast-cancer survivors did not improve torque or 10RM strength, and 7 days after stroke did not improve handgrip or muscle mass indices.182730 That does not contradict the broader positive story. It suggests creatine works best when there is enough training stimulus, enough time, or both, rather than acting like an acute performance switch for every population.12162937
The broad heterogeneity also has a likely methodological explanation. Studies mixed loading protocols with steady daily dosing, used populations ranging from elite athletes to frail older adults, and measured outcomes with everything from handgrip dynamometers to 1RM lifts to cycle ergometers. When an evidence base is that diverse, prediction intervals crossing no effect are not a sign that the average benefit is fake. They are a sign that real-world response depends on context.41217202632 The current analysis cannot cleanly separate who benefits most, but the recurring winners are people doing resistance or high-intensity training and outcomes that directly reflect repeated force production.
Discussion
The overall conclusion is confident but not indiscriminate: creatine shows that it improves adult strength and several related muscle-performance outcomes, and it is one of the stronger evidence bases in supplement research.312162937 The highest-confidence claims in this review are about strength-related performance and body-composition support, not about every possible metric of athleticism.
What is supported well is straightforward. Creatine improves overall strength on average, lower-body force production looks especially responsive, peak anaerobic power trends positive, lean mass moves upward, and recovery after hard exercise appears better.312161729323637 The benefits are usually modest. Most people should think in terms of slightly better training adaptations, a bit more work capacity in short hard efforts, and somewhat better recovery, not dramatic overnight transformation.
What is not well supported is just as important. The current analysis does not show a dependable jump-performance benefit, does not firmly establish a specific upper-body pressing advantage, and does not justify a strong general claim that creatine reduces fatigue in all contexts.11172332 Several respectable studies also found no meaningful effect in COPD, PAD, immobilization, stroke recovery, or very short standalone dosing windows, which keeps the review grounded.4142130
What would change confidence from here is not another tiny short-term trial in young men. The most useful next studies would directly compare loading versus no loading, separate trained from untrained participants, measure muscle creatine uptake, and use standardized strength and body-composition outcomes over meaningful training periods. More female-specific and clinical-population research would also matter, because the current literature is still disproportionately built around male or mixed groups and resistance-training contexts.24283233
Taken together, the evidence reviewed here supports creatine as a useful but not magical supplement. If the goal is to improve strength, support lean-mass gains, and recover a bit better from hard training, the case is strong. If the goal is a guaranteed increase in jump height, universal anti-fatigue effects, or large body-composition changes without serious training, the evidence is not there.
Methodology
We searched PubMed for studies on creatine and strength or muscle performance, then filtered to controlled human studies that matched the review question. The final review included 37 studies, mostly randomized placebo-controlled trials in adults, covering resistance training, sport performance, rehabilitation, and several clinical populations.
We read each study and recorded what it measured, how large it was, how long it lasted, and what it found. We assessed evidence quality with the GRADE framework and judged clinical importance against published meaningful-change thresholds when those were available.
A note of caution: GRADE was designed for pharmaceutical interventions and often rates nutrition and supplement evidence conservatively. It automatically downgrades observational evidence and only gives special credit for very large effects, so supplement literatures can look "low certainty" even when multiple randomized trials point in the same direction. Our 0 to 100 trust scoring adds a more continuous read on how likely an effect is to be both real and meaningful. When those two systems diverge, the difference usually reflects GRADE's structural conservatism rather than hidden evidence against the supplement.
Every cited study is publicly indexed on PubMed. Known limitations include heterogeneous populations and outcome measures, incomplete subgroup data on sex and training status, variable dosing strategies, and some body-composition methods that can blur true muscle gain with changes in body water.
Study Selection
Characteristics of Included Studies
| Study | Design | N | Population | Dose | Duration | RoB |
|---|---|---|---|---|---|---|
| K 1997 FT | rct | 19 | healthy | 20 g/day for 4 days | 4 days high-dose phase followed by 10 wk resistance training with low-dose creatine; subgroup additionally underwent 10 wk detraining with low-dose creatine and then 4 wk after cessation | Low |
| M 1999 FT | rct | 42 | NR | NR | NR | Some |
| D 2001 FT | rct | ? | NR | 5-day loading then 5-week maintenance (0.3 to 0.03 g/kg/day dosing per paper) | 6 weeks | Some |
| A 2003 FT | rct | 28 | healthy | 5 g creatine daily for 14 weeks | 14 weeks | Some |
| G 2006 FT | rct | 23 | clinical | 0.3 g/kg/day for 1 week then 0.07 g/kg/day for 7 weeks | 8 weeks | Some |
| G 2009 FT | rct | 40 | clinical | 20 g/day for 5 days then 4.8 g/day (maintenance) for 14 weeks | 14 weeks | Some |
| Y 2009 FT | rct | 17 | healthy | 20 g/day for 5 days | 5-day supplementation period (baseline at day 0; performance testing on day 3 and day 6) | Some |
| J 2012 FT | rct | 22 | healthy | 20 g/day for 7 days | 7 days | Some |
| C 2013 FT | rct | 56 | healthy | 20 g/day for 5 days, then 5 g/day for 24 weeks | 24 weeks | Some |
| C 2014 FT | rct | 77 | healthy | 1.25 g/day for 28 days | 28 days | Low |
| B 2014 FT | rct | 74 | clinical | Creatine 5 g/day (with a 20 g/day 5‑day loading phase) | 24 weeks | Some |
| J 2014 FT | rct | 23 | healthy | Creatine 20 g/day for 1 week, then 5 g/day for 6 more weeks | 7 weeks (pre-season) | High |
| D 2015 FT | rct | 64 | healthy | 0.1 g/kg immediately after workouts (on training days) | 32 weeks | High |
| E 2016 FT | rct | 54 | healthy | N/A (placebo) | 5 days supplementation, testing on day 6 | Some |
| E 2017 FT | rct | 30 | healthy | 20 g/day for 5 days (loading), then 5 g/day maintenance | Approximately 19 days (5-day loading, 7-day immobilization, 7-day recovery) | Some |
| S 2018 FT | rct | 14 | clinical | 3 g creatine daily for 8 weeks | 8 weeks | Some |
| C 2018 FT | rct | 30 | healthy | 20 g/day for 6 days, then 2 g/day during 4 weeks | 4 weeks (6-day loading phase followed by 4-week maintenance during training) | Some |
| M 2019 FT | rct | 18 | healthy | 70.0mg/kg/day | 8 weeks | Low |
| J 2019 FT | rct | 33 | healthy | 0.3 g/kg/day for 10 days | 10 days | Some |
| C 2019 FT | rct | 56 | clinical | 20 g/day for 5 days (loading), then 5 g/day maintenance | 12-week intervention (plus 2-week familiarisation and 2-week post-testing; total ≈16 weeks) | Low |
| D 2021 FT | rct | 46 | NR | Creatine 0.1 g/kg/day for 12 months (with resistance training) | 12 months | Some |
| W 2021 FT | rct | 32 | clinical | 20 g/day for 1 week, then 5 g/day for 7 weeks | 8 weeks | Some |
| J 2021 FT | rct | 11 | healthy | ≈26.2 g daily for 7 days (divided into 4 servings) | 7 days (per intervention), crossover with ≥30 day washout | Low |
| M 2022 FT | rct | 20 | clinical | 0.3 g/kg/day creatine loading for 7 days | 28 days (4 weeks of beta-alanine; creatine loading added during the final 7 days) | Some |
| S 2023 FT | rct | 13 | healthy | 0.1 g/kg/day creatine for 6 weeks | 42 days | Some |
| A 2024 FT | rct | 40 | clinical | 5 g creatine daily for 6 months | 12 months | Some |
| D 2024 FT | rct | 23 | healthy | 20 g once daily for 7 days | 6-day training camp (7 days supplementation, including day before camp) | Low |
| E 2024 FT | rct | 19 | clinical | 20 g/day (5 g four times daily) for 7 days | 7 days (two test sessions separated by 7 days; supplementation between sessions) | Low |
| A 2024 FT | rct | 25 | healthy | 5 g creatine daily for 6 weeks | 6 weeks | Some |
| E 2024 FT | rct | 40 | healthy | 0.03 g/kg/day creatine HCl with RT for 8 weeks | 8 weeks | Some |
| J 2024 FT | rct | 30 | clinical | 20 g/day for 7 days | 7-day intervention with 90-day follow-up | Low |
| B 2025 FT | crossover trial | 25 | healthy | 20 g/day (4 × 5 g) for 5 days | 3 weeks (four visits within a 3-week period); creatine/placebo administered for 5 days before the respective trial | Some |
| S 2025 FT | rct | 40 | healthy | 3 g daily for 33 days (28 pre-exercise, 5 post) | 33 days (28 days pre-exercise supplementation, 5 days post-exercise) | High |
| T 2025 FT | rct | 15 | healthy | 0.3 g/kg/day (divided QID) for 7 days | 7 days supplementation; post-testing 24–48 h after final dose | Low |
| M 2025 FT | rct | 40 | clinical | Creatine: 20g/day loading (week 1), then 5g/day for 3 more weeks | 4 weeks | Low |
| K 2025 FT | rct | 14 | healthy | 20 g/day (4 × 5 g) for 7 days | >28 days (includes familiarization, two 7-day intervention periods, and a 14-day washout) | Low |
| A 2025 FT | rct | 10 | healthy | 0.3 g/kg/day (≈300 mg/kg/day) for 3 days, divided dosing | Each supplementation arm: 3 days (with first dose 2 h pre-test); crossover with 7-day washout; entire study ~2 weeks | Low |
Sources
- 1. M 1999. Effects of in-season (5 weeks) creatine and pyruvate supplementation on anaerobic performance and body composition in American football players. (1999) ↑
- 2. D 2001. Effects of creatine monohydrate supplementation during combined strength and high intensity rowing training on performance. (2001) ↑
- 3. A 2003. Creatine supplementation enhances isometric strength and body composition improvements following strength exercise training in older adults. (2003) ↑
- 4. G 2006. Creatine supplementation and physical training in patients with COPD: a double blind, placebo-controlled study. (2006) ↑
- 5. G 2009. Creatine fails to augment the benefits from resistance training in patients with HIV infection: a randomized, double-blind, placebo-controlled study. (2009) ↑
- 6. Y 2009. Effects of two and five days of creatine loading on muscular strength and anaerobic power in trained athletes. (2009) ↑
- 7. J 2012. The effects of creatine monohydrate loading on anaerobic performance and one-repetition maximum strength. (2012) ↑
- 8. C 2014. The effects of polyethylene glycosylated creatine supplementation on anaerobic performance measures and body composition. (2014) ↑
- 9. C 2013. Creatine supplementation associated or not with strength training upon emotional and cognitive measures in older women: a randomized double-blind study. (2013) ↑
- 10. B 2014. Creatine supplementation and resistance training in vulnerable older women: a randomized double-blind placebo-controlled clinical trial. (2014) ↑
- 11. J 2014. Creatine monohydrate supplementation on lower-limb muscle power in Brazilian elite soccer players. (2014) ↑
- 12. D 2015. Strategic creatine supplementation and resistance training in healthy older adults. (2015) ↑
- 13. E 2016. Effects of Coffee and Caffeine Anhydrous Intake During Creatine Loading. (2016) ↑
- 14. E 2017. Creatine Loading Does Not Preserve Muscle Mass or Strength During Leg Immobilization in Healthy, Young Males: A Randomized Controlled Trial. (2017) ↑
- 15. S 2018. Creatine or vitamin D supplementation in individuals with a spinal cord injury undergoing resistance training: A double-blinded, randomized pilot trial. (2018) ↑
- 16. M 2019. Creatine monohydrate supplementation during eight weeks of progressive resistance training increases strength in as little as two weeks without reducing markers of muscle damage. (2019) ↑
- 17. C 2018. Effects of 4-Week Creatine Supplementation Combined with Complex Training on Muscle Damage and Sport Performance. (2018) ↑
- 18. J 2019. Effect of Creatine Supplementation Dosing Strategies on Aging Muscle Performance. (2019) ↑
- 19. C 2019. Examining the effects of creatine supplementation in augmenting adaptations to resistance training in patients with prostate cancer undergoing androgen deprivation therapy: a randomised, double-blind, placebo-controlled trial. (2019) ↑
- 20. D 2021. Effect of 12 months of creatine supplementation and whole-body resistance training on measures of bone, muscle and strength in older males. (2021) ↑
- 21. W 2021. Effect of Creatine Supplementation on Functional Capacity and Muscle Oxygen Saturation in Patients with Symptomatic Peripheral Arterial Disease: A Pilot Study of a Randomized, Double-Blind Placebo-Controlled Clinical Trial. (2021) ↑
- 22. J 2021. Short-Term Creatine Loading Improves Total Work and Repetitions to Failure but Not Load-Velocity Characteristics in Strength-Trained Men. (2021) ↑
- 23. M 2022. Effects of Four Weeks of Beta-Alanine Supplementation Combined with One Week of Creatine Loading on Physical and Cognitive Performance in Military Personnel. (2022) ↑
- 24. S 2023. Creatine monohydrate supplementation changes total body water and DXA lean mass estimates in female collegiate dancers. (2023) ↑
- 25. A 2024. Effect of Creatine Supplementation on Body Composition and Malnutrition-Inflammation Score in Hemodialysis Patients: An Exploratory 1-Year, Balanced, Double-Blind Design. (2024) ↑
- 26. D 2024. High-dose short-term creatine supplementation without beneficial effects in professional cyclists: a randomized controlled trial. (2024) ↑
- 27. E 2024. Impact of Short-Term Creatine Supplementation on Muscular Performance among Breast Cancer Survivors. (2024) ↑
- 28. A 2024. Creatine Improves Total Sleep Duration Following Resistance Training Days versus Non-Resistance Training Days among Naturally Menstruating Females. (2024) ↑
- 29. E 2024. Supplementing With Which Form of Creatine (Hydrochloride or Monohydrate) Alongside Resistance Training Can Have More Impacts on Anabolic/Catabolic Hormones, Strength and Body Composition? (2024) ↑
- 30. J 2024. Influence of CReatine Supplementation on mUScle Mass and Strength After Stroke (ICaRUS Stroke Trial): A Randomized Controlled Trial. (2024) ↑
- 31. B 2025. Supplementation of Creatine Monohydrate Improves Sprint Performance but Has no Effect on Glycolytic Contribution: A Nonrandomized, Placebo-Controlled Crossover Trial in Trained Cyclists. (2025) ↑
- 32. S 2025. The Effects of Creatine Monohydrate Supplementation on Recovery from Eccentric Exercise-Induced Muscle Damage: A Double-Blind, Randomized, Placebo-Controlled Trial Considering Sex and Age Differences. (2025) ↑
- 33. T 2025. Muscle creatine levels and sprint performance in young adult vegans and vegetarians after 7 days of creatine monohydrate supplementation. (2025) ↑
- 34. M 2025. Additional Benefits of Creatine Supplementation with Physical Therapy and Resistance Exercise in Knee Osteoarthritis: A Randomized Controlled Trial. (2025) ↑
- 35. K 2025. Effects of Creatine Monohydrate Loading on Sleep Metrics, Physical Performance, Cognitive Function, and Recovery in Physically Active Men: A Randomized, Double-Blind, Placebo-Controlled, Crossover Trial. (2025) ↑
- 36. A 2025. Short-term creatine supplementation enhances strength, reduces fatigue, and accelerates recovery in resistance-trained athletes: a double-blind, randomized, crossover trial. (2025) ↑
- 37. K 1997. Long-term creatine intake is beneficial to muscle performance during resistance training. (1997) ↑