Caffeine and Exercise Performance

Sep 2

5 min read

Caffeine was a banned substance in sport for most of forty years. It is now the most thoroughly studied legal performance aid in existence, with twenty-one meta-analyses pooled into one umbrella review. The evidence is unusually clean: it works, the effect is small to moderate, it works across almost every exercise type, and the dose that works is lower than most people take.

The evidence

Pesta and colleagues (2013) reviewed caffeine's mechanisms - adenosine-receptor blockade, increased fat oxidation, calcium handling - and its regulatory history: on the banned list from 1962 to 1972 and again from 1984 to 2003, with a urinary threshold of 12 micrograms per millilitre counting as doping. Bishop (2010) reviewed supplements for team sport and reported good evidence that caffeine improves single- and multiple-sprint performance. Neither paper pools numbers; the meta-analyses that followed do.

What happened

Umbrella review (Grgic and colleagues, 2020; 21 meta-analyses): standardised effects of 0.22 to 0.61 for aerobic endurance, 0.16 to 0.20 for muscle strength, 0.28 to 0.38 for muscular endurance, 0.18 to 0.27 for anaerobic power, 0.17 for vertical jump and 0.41 for exercise speed. Evidence graded moderate; most common dose 6 mg/kg; 72 to 100% of participants male.
Endurance (Southward and colleagues, 2018; 46 studies): mean power output +3.03% (ES 0.23) and time-trial completion 2.22% faster (ES 0.41) at 3 to 6 mg/kg. Seven studies reported individuals who performed worse.
Strength and power (Grgic and colleagues, 2018; 149 and 145 participants): strength SMD 0.20 (95% CI 0.03 to 0.36, p = 0.023); power SMD 0.17 (0.00 to 0.34, p = 0.047); upper body 0.21 (0.02 to 0.39) versus lower body 0.15 (-0.05 to 0.34, not significant). PEDro quality 8 to 10 out of 10.
Position stand (Guest and colleagues, 2021): 3 to 6 mg/kg is the consistently effective range; 2 mg/kg may be the minimum; 60 minutes pre-exercise is the usual timing; gum, gels and mouth rinses also work; response varies with genetics (CYP1A2, ADORA2A) and habitual intake.

The current answer

Xue and colleagues (2025) ran a network meta-analysis of 48 time-trial studies and 612 participants (89% male) comparing doses and delivery methods. Low-dose capsules of about 3 mg/kg ranked first: completion time SMD -0.34 (95% CI -0.62 to -0.06), roughly 2.2% faster than placebo; moderate-dose capsules (4 to 6 mg/kg) SMD -0.31 (-0.45 to -0.17); moderate-dose gum -0.30 (-0.57 to -0.02). No trial was at high risk of bias. Xiao and colleagues (2025; 12 crossover trials, 230 resistance-trained participants) found caffeine at 3 to 12 mg/kg improved mean lifting velocity (SMD 0.42, 0.19 to 0.65) and mean power (0.21, 0.12 to 0.30), with much larger effects in low habitual users (SMD 0.87 vs 0.21, p < 0.01). Yu and colleagues (2025; 13 studies) found repeated-sprint peak power rose 5.3 W (2.5 to 8.1), significant at 6 mg/kg or more but not below.

On status: caffeine was removed from the prohibited list in 2004 and sits on the 2026 WADA Monitoring Programme for in-competition use - monitored, not banned - confirmed by the Athletics Integrity Unit and USADA. The historical 12 micrograms per millilitre threshold no longer applies; the NCAA retains a 15 micrograms per millilitre limit, far above ergogenic doses.

How certain is this?

Certainty: high that caffeine improves endurance, strength, power and sprint performance by small-to-moderate amounts; moderate on the exact size per outcome, because the primary trials are small crossovers and heterogeneity is high for some measures. Umbrella-review quality scores ranged from 44 to 88%.

Who this covers: overwhelmingly young men - 72 to 100% male across the meta-analyses. Effects in women appear in the same direction but smaller in some analyses (repeated-sprint power +4.0 W vs +13.1 W). Habitual heavy users show smaller acute gains in the newest data; older adults and adolescents are barely studied.

What would change our mind: a large trial in trained women comparing 3 and 6 mg/kg with placebo on a time trial, stratified by habitual intake and CYP1A2 genotype. If the female effect vanished, the population claim would narrow.

Caffeine works for almost everything, by a few percent, and the dose that does it is about 3 mg/kg an hour before - not the double espresso and a pre-workout stacked on top.

What this means in practice

The trials used 3 to 6 mg/kg roughly 60 minutes before - for a 75 kg athlete that is 225 to 450 mg, and the 2025 network analysis found the bottom of that range ranked best for time trials.
Expect about 2 to 3% on endurance work and effect sizes of 0.2 on strength - the difference between places at the sharp end, not a new athlete.
It is legal. Monitored by WADA in competition, not prohibited; no urinary threshold in WADA sport since 2004.
Habit blunts it. In the 2025 resistance-training analysis, low habitual users (under 3 mg/kg/day) gained four times the effect of regular users.
Some people get worse. Seven of 46 endurance trials reported negative responders; genetics and gut tolerance are why the trials say "test it in training first".
Repeated sprints needed the higher end: 6 mg/kg or more was significant, below it was not, in the 2025 analysis - one of the few outcomes where more was better.

Want this applied to your own programme, your sport and your numbers? Book a Private Consultation - a 60-minute direction session with Coach Ohana.

References

Grgic J, Grgic I, Pickering C, Schoenfeld BJ, Bishop DJ, Pedisic Z. Wake up and smell the coffee: caffeine supplementation and exercise performance - an umbrella review of 21 published meta-analyses. Br J Sports Med. 2020;54(11):681-688. doi:10.1136/bjsports-2018-100278

Xue R, Huang J, Chen B, et al. Effects of caffeine dose and administration method on time-trial performance: a systematic review and network meta-analysis. Nutrients. 2025;17(23):3792. doi:10.3390/nu17233792

Southward K, Rutherfurd-Markwick KJ, Ali A. The effect of acute caffeine ingestion on endurance performance: a systematic review and meta-analysis. Sports Med. 2018;48(8):1913-1928. doi:10.1007/s40279-018-0939-8

Grgic J, Trexler ET, Lazinica B, Pedisic Z. Effects of caffeine intake on muscle strength and power: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2018;15:11. doi:10.1186/s12970-018-0216-0

Guest NS, VanDusseldorp TA, Nelson MT, et al. International Society of Sports Nutrition position stand: caffeine and exercise performance. J Int Soc Sports Nutr. 2021;18:1. doi:10.1186/s12970-020-00383-4

Xiao Y, Ding L, Xu Z, et al. Effects of acute caffeine intake on muscular power during resistance exercise: a systematic review and meta-analysis. Front Nutr. 2025;12:1686283. doi:10.3389/fnut.2025.1686283

Yu L, Wang Y, Su W, et al. Effects of acute caffeine ingestion on repeated sprint ability: a systematic review and meta-analysis. Nutrients. 2025;17(21):3475. doi:10.3390/nu17213475

Pesta DH, Angadi SS, Burtscher M, Roberts CK. The effects of caffeine, nicotine, ethanol, and tetrahydrocannabinol on exercise performance. Nutr Metab (Lond). 2013;10(1):71. doi:10.1186/1743-7075-10-71

Bishop D. Dietary supplements and team-sport performance. Sports Med. 2010;40(12):995-1017. doi:10.2165/11536870-000000000-00000

Last reviewed: 2 September 2026. Re-review on the next caffeine meta-analysis, any change to the WADA Monitoring Programme, or by September 2028.