Supplements for athletes: what the evidence supports for physical performance

This hub groups four outcomes, and each grade belongs to one compound and one outcome. A result for endurance says nothing about strength, so the page keeps them apart.

Strength and power output covers maximal lifts such as the one-repetition maximum (1-RM), the heaviest weight lifted once [3], isometric force tests such as the mid-thigh pull on a force platform [5], and maximal muscle power measured in a rested, non-fatigued state [6].

Endurance performance covers time to exhaustion at a fixed intensity, self-paced time trials and maximal oxygen uptake (VO2max), the most oxygen the body can use during hard exercise [7]. These tests do not always agree: in a beta-alanine meta-analysis, effects in exercise lasting 0.5-10 minutes were larger for capacity tests (how long you can keep going) than for performance tests (how fast you finish a set task) [8].

Lean mass is fat-free mass, the body's weight minus fat, which trials measure with body scans such as dual-energy X-ray absorptiometry (DXA) [9][3].

Reaction time comes from computer tasks such as choice reaction tests [10]. It also appears on our focus hub.

Several results below are standardized effect sizes (g, d, ES or SMD). They put different tests on one scale: values near zero mean little or no difference from the comparison, and larger values mean bigger differences.

Most participants in the caffeine meta-analyses were young men [11], and the training and sleep analyses below were mostly male [2][4], so results may not apply equally to women and older adults.

Supplements sit on top of training, food and sleep, and the IOC statement says supplements can make a minor contribution to an athlete's nutrition program [1].

Training. A 2026 meta-regression of 67 studies with 2,058 participants (79.1% male) found that gains in muscle size and strength rose with weekly set volume, with diminishing returns that were more pronounced for strength [2]. Training a muscle more often each week was linked to larger strength gains, while its effect on muscle size was compatible with negligible effects [2].

Protein. In a meta-analysis of 49 randomized trials (1,863 healthy adults) of resistance training lasting at least 6 weeks, protein supplementation increased 1-RM strength gains by 2.49 kg and fat-free mass gains by 0.30 kg compared with control conditions [3]. Total protein intakes above 1.62 g/kg/day brought no further fat-free mass gain [3]. One author reports grants and honoraria from the US National Dairy Council, which supported trials in the analysis, and a 2020 correction added another author's advisory role with a sports supplement maker [3].

Sleep. A meta-analysis of 69 publications found that sleep loss (6 hours or less in 24 hours) lowered exercise performance by a mean of 7.56% compared with normal sleep, with effects in every exercise category studied; 89% of participants were male [4]. Results varied widely between studies, and consistent effects appeared only with full sleep deprivation and with waking earlier than usual [4]. A systematic review of 25 sleep intervention studies in athletes found sleep extension and naps the most effective strategies, while noting the limited amount of high-quality evidence [12].

Planning. The IOC statement advises a complete nutritional assessment before decisions about supplements, and trialling any performance supplement in training before using it in competition [1].

Contamination. For tested athletes, the IOC statement describes inadvertent ingestion of substances prohibited under anti-doping codes as a known risk of some supplements [1]. A 2015 review cited data suggesting that 10-15% of supplements may contain prohibited substances, while concluding that the scale of the problem cannot yet be quantified [13]. A 2023 review reported that analytical studies found prohibited substances in 14 to 50% of supplement samples, and listed third-party certification among safer practices; several of its authors have consulted for or received funding from companies that sell dietary ingredients [14].

The ranked list on this page renders from the editor-approved grades. Each summary below names the population and the comparison behind the grade. Each monograph linked below has the compound's safety section and known interactions.

Caffeine

An umbrella review of 21 meta-analyses found caffeine ergogenic for aerobic endurance, muscle strength, muscle endurance, power, jumping and speed, with evidence mostly rated moderate quality [11]. Not every analysis gave a definite direction once prediction intervals were considered, and most participants were young men [11]. In rested healthy adults, a meta-analysis of 31 placebo-controlled trials (1,455 participants) found faster reaction time (g = 0.28) [15]. Read the full caffeine monograph.

Creatine

In a three-level meta-analysis of 63 studies (1,489 participants), creatine plus resistance training increased upper-body strength (g = 0.54), lower-body strength (g = 0.55) and lean mass (g = 0.30) compared with training alone, with no significant change in fat mass [16]. Strength effects were larger in younger participants, an exploratory finding [16]. In older adults (mean age about 64), a meta-analysis of placebo-controlled trials found greater fat-free mass and chest press and leg press gains than training alone [17]. See creatine and our guide to creatine for women.

Beetroot nitrate

A meta-analysis of 80 placebo-controlled crossover studies, funded by the US National Institutes of Health, found a small overall improvement in exercise performance in young, healthy adults (d = 0.174) and a clear effect in recreationally active young men, but no effect was observed in well-trained endurance athletes or in studies of women only [18]. An individual participant data meta-analysis of 19 double-blind crossover studies (268 participants) found higher maximal muscle power (g = 0.42) [6]. In a 2026 meta-analysis of 45 trials in 659 athletes, no pooled effect was statistically significant [19]. See beetroot nitrate.

Beta-alanine

A meta-analysis of 40 double-blind, placebo-controlled studies (1,461 participants) found a small overall effect (0.18), larger for exercise-capacity tests than for performance tests [8]. A 2026 preregistered meta-analysis of 17 randomized trials found no statistically significant improvement in repeated sprint ability [20], and a meta-analysis of 20 trials (492 participants) found no significant effect on fat-free mass [21]. One author of the body-composition analysis has conducted industry-sponsored research on nutraceuticals [21]. See beta-alanine.

Rhodiola rosea

A 2025 meta-analysis of 26 randomized trials in 668 healthy participants (mean age 22.0) found improvements in VO2max (ES = 0.32), time to exhaustion and time-trial performance, with heterogeneity the authors say warrants caution [7]. A 2026 meta-analysis of 9 trials across Rhodiola species found no significant effect on VO2max [22]. See Rhodiola rosea.

Ashwagandha

A 2026 three-level meta-analysis of 13 trials (599 participants, mainly young adults or athletes) found the most consistent benefit for aerobic endurance (g = 0.54), but the prediction interval for overall performance crossed zero [23]. In resistance-training trials, a 2026 meta-analysis of 3 poolable trials found greater 1-RM gains than placebo, but both strength estimates lost significance under a stricter adjustment and certainty was rated low [24]. All three pooled trials were funded or supplied by ashwagandha manufacturers [24]. See ashwagandha.

Taurine

A 2018 meta-analysis of studies of isolated oral taurine found improved endurance performance (Hedges' g = 0.40), with doses of 1 to 6 g/day given once or for up to 2 weeks [25]. A 2025 meta-analysis of 23 randomized single-dose trials (308 participants) found that a single dose may be associated with a small improvement in overall exercise performance (g = 0.25); prediction intervals often included no effect, and certainty was rated low to very low [26]. Reaction-time findings come mostly from taurine combined with caffeine: a network meta-analysis of 12 studies found faster reaction time with the combination than with either alone [27]. In a double-blind trial of 48 habitual caffeine users that tested each ingredient separately, caffeine shortened choice reaction time while taurine alone made it slower; the authors concluded that caffeine, not taurine, is likely responsible for the cognitive changes reported after energy drinks [37]. The direction for taurine itself is therefore not consistent. See taurine.

L-theanine

In a 2026 meta-analysis of 31 randomized trials in healthy and clinical populations, a single 200 mg dose taken 30-60 minutes before cognitive testing improved choice reaction time (SMD = 0.51); one author founded a company selling dietary supplements [10]. These were cognitive tests, not sport tasks [10]. See L-theanine.

Alpha-GPC

In a double-blind crossover trial in 13 college-aged men, funded by the ingredient supplier Chemi Nutra, 6 days of 600 mg a day increased mid-thigh pull peak force more than placebo, while upper-body force did not differ significantly [5]. In an acute crossover trial in 20 resistance-trained men, funded by the ingredient supplier NNB Nutrition, single doses changed Stroop scores but not physical performance [28]. See alpha-GPC.

Phosphatidylserine

In one double-blind trial in 14 active men, 750 mg a day of soybean-derived phosphatidylserine for 10 days lengthened time to exhaustion at 85% VO2max, with no change on placebo; the authors describe it as the first such report [29]. See phosphatidylserine.

Schisandra

In a double-blind trial of 54 adults over 50 with relatively low muscle mass who walked regularly, 1 g a day of Schisandra chinensis extract for 12 weeks increased right knee extensor strength by 10.2 Nm more than placebo, with no difference in muscle mass [30]. See schisandra.

PQQ

In 23 untrained men doing 6 weeks of supervised endurance training, 20 mg/day of pyrroloquinoline quinone (PQQ) produced no significant difference from placebo in aerobic performance [9]. See PQQ.

Lion's mane

In a placebo-controlled crossover trial in healthy adults aged 18 to 50, funded in part by the ingredient supplier Applied Food Sciences, reaction time on some tasks improved after a single 1 g dose of lion's mane [31]. A 4-week trial of 10 g a day in 24 college-age adults found no effect on cognition [32]. See lion's mane.

Magnesium

A meta-analysis of 14 randomized trials found no significant improvement in muscle strength or power, with no apparent benefit in athletes and physically active people; effects appeared more pronounced in older adults and people with alcoholism, a clinical population [33]. See magnesium.

Panax ginseng

In a meta-analysis of 8 placebo-controlled trials, ginseng was not associated with better physical performance (SMD = -0.01) [34]. A 2022 meta-analysis of 14 studies in athletes and active people found a significant short-term effect only for vertical jump at high doses, and no effect on VO2max or peak power output [35]. See Panax ginseng.

L-tyrosine

A 2023 meta-analysis of 10 interventions from 8 studies in physically active people found no significant difference from placebo for time to exhaustion (SMD = 0.02) or time-trial performance, with moderate-quality evidence [36]. See L-tyrosine.

Our guide to creatine for women covers the population that most trials on this page under-sample. The evidence matrix shows every graded compound and outcome side by side, and how we grade evidence explains each grade and modifier.

Several syntheses published in 2026 changed or tested the picture. A meta-analysis of 45 trials in athletes found no significant pooled effect of dietary nitrate [19]. A preregistered meta-analysis found no statistically significant effect of beta-alanine on repeated sprints [20]. A resistance-training meta-analysis found ashwagandha's strength estimates statistically fragile [24], and a meta-analysis of 63 studies found strength and lean-mass gains when creatine was added to training [16].

Which supplements have the best evidence for athletes?

An IOC consensus statement names caffeine, creatine, specific buffering agents and nitrate among the few performance supplements with good evidence, and notes that responses vary between individuals [1]. Each compound's grade on this page is set per outcome.

Do supplements matter more than training, protein and sleep?

The IOC statement says supplements can make a minor contribution to an athlete's nutrition program [1]. Training volume, protein intake and sleep each showed measurable effects on strength, muscle or performance in meta-analyses [2][3][4].

Can a supplement cause a failed drug test?

The IOC statement describes inadvertent ingestion of prohibited substances as a known risk of some supplements [1]. A 2023 review reported prohibited substances in 14 to 50% of samples in analytical studies and listed third-party certification among safer practices; several of its authors have consulted for companies that sell dietary ingredients [14].

Does beetroot juice help trained athletes?

An NIH-funded meta-analysis found a small overall benefit in young, healthy adults and a clear effect in recreationally active young men, but no effect was observed in well-trained endurance athletes [18]. A 2026 meta-analysis of 45 trials in athletes found no statistically significant pooled effect [19].