The short version
- The ISSN position stand reports caffeine consistently improves exercise performance at 3 to 6 mg per kilogram of body mass, with doses around 9 mg/kg producing side effects without added benefit.
- In a meta-analysis of 46 studies, caffeine at 3 to 6 mg/kg improved mean power output by 3.03% and time-trial completion time by 2.22%, though two studies reported slower time trials and five reported lower power.
- Caffeine's effect on strength is small: a meta-analysis of ten studies found a standardised mean difference of 0.20 for strength and 0.17 for power, significant for upper body strength but not lower body.
- In 101 competitive male athletes, 4 mg/kg of caffeine improved 10 km cycling time by 6.8% in those with the CYP1A2 AA genotype, had no effect in AC, and made CC athletes 13.7% slower.
- An umbrella review of 21 meta-analyses found caffeine ergogenic across endurance, strength, muscle endurance, power, jumping and speed, with evidence quality generally graded moderate and most studies conducted in young men.
Very few supplements survive an umbrella review. Caffeine does. Grgic and colleagues pooled 11 published reviews containing 21 separate meta-analyses and found caffeine ergogenic for aerobic endurance, muscle strength, muscle endurance, power, jumping performance and exercise speed. The quality of evidence was generally graded moderate, with some low and very low, and they noted that most primary studies were conducted in young men.
That is the most robust position any ingredient in this category holds. The interesting part is not whether it works. It is that the dose is expressed in milligrams per kilogram of bodyweight, and almost nobody takes it that way.
The number to actually use
The ISSN position stand is specific: caffeine has consistently been shown to improve exercise performance at doses of 3 to 6 mg per kilogram of body mass. The minimal effective dose remains unclear but may be as low as 2 mg/kg. Very high doses around 9 mg/kg are associated with a high incidence of side effects and do not appear to be required for an ergogenic effect. The most commonly used timing is 60 minutes before exercise, though the optimal timing depends on the source — caffeine chewing gum, for instance, needs a shorter wait than capsules.
For a 60 kg adult that range is 180 to 360 mg. For a 95 kg adult it is 285 to 570 mg. The same scoop of pre-workout supplement is a different intervention for those two people, which is one of the reasons individual experiences with the same tub diverge so wildly.
How big the effect is, by activity
Endurance is where it looks best. Southward and colleagues meta-analysed 46 studies and found that at moderate doses of 3 to 6 mg/kg, caffeine improved mean power output by 3.03% and time-trial completion time by 2.22% against placebo. They also reported, plainly, that two studies found slower time-trial performance and five found lower mean power output with caffeine — the average is not the whole distribution.
Strength and power are more modest. Grgic and colleagues meta-analysed ten studies on each and found significant but small effects: strength SMD 0.20 (p = 0.023) and power SMD 0.17 (p = 0.047). The subgroup analysis is the honest detail — the effect was significant for upper body strength (0.21) but not for lower body (0.15, p = 0.147). The authors also recommended future studies control blinding more rigorously, which is a polite way of noting that people can usually tell whether they have had caffeine.
The genotype result nobody quotes
This is the finding that should stop anyone from assuming more is better. Guest and colleagues gave 101 competitive male athletes 0, 2 or 4 mg/kg before a 10 km cycling time trial, and genotyped them for a variant in the CYP1A2 gene affecting caffeine metabolism.
The overall result looked routine: 4 mg/kg cut cycling time by 3%. Then the interaction. Among athletes with the AA genotype, time fell 4.8% at 2 mg/kg and 6.8% at 4 mg/kg. Among those with the AC genotype, there was no effect at either dose. And among those with the CC genotype, 4 mg/kg increased cycling time by 13.7% — they were substantially slower on caffeine than on placebo.
One study, 101 male athletes, one distance. But it puts a number on something most people have already noticed anecdotally: a population-average benefit is not a promise to any individual. The umbrella review makes the same point more formally, noting that not all analyses provided a definite direction for the effect of caffeine once the 95% prediction interval was considered.
What we'd actually tell you
Work out your own number: bodyweight in kilograms times three, in milligrams, taken about 60 minutes before a session that matters. That is the low end of the effective range and the sensible place to start. Going to six is available if three does nothing; going to nine buys side effects rather than performance.
Then test it honestly. Use it on hard sessions and compare against sessions without it, on something measurable — a time, a rep count, a distance — rather than on how alert you felt. If it does nothing twice at a properly calculated dose, you may simply be someone it does not work for, and the genotype data says that is a real category rather than a failure of effort.
Two caveats we would rather say ourselves. Daily use changes the question entirely — at that point much of what you feel from a morning dose is your own baseline being restored, which is a separate article and a separate problem. And caffeine is the one ingredient in a pre-workout label that is doing most of the visible work; beta-alanine will make you tingle, but the alertness, the willingness and the extra rep are almost all this.
Good questions
How much caffeine should I take before training?
Three to six milligrams per kilogram of bodyweight, about 60 minutes before. For a 70 kg adult that is 210 to 420 mg. Start at the bottom of the range: the position stand notes the minimal effective dose may be as low as 2 mg/kg, and that around 9 mg/kg brings a high incidence of side effects without added performance. The dose is per kilogram of you, not per scoop.
Is more caffeine better?
No, and past a point it is worse. Doses around 9 mg/kg are associated with a high rate of side effects and are not required for the ergogenic effect. In one trial, athletes with a particular CYP1A2 genotype were 13.7% slower over 10 km on 4 mg/kg than on placebo. If your current dose does nothing, the answer is more likely to be that you are a non-responder than that you need double.
Does caffeine actually help with lifting, or just cardio?
Both, but by clearly different amounts. Endurance shows the most consistent moderate benefit, with 3.03% more mean power output in a 46-study meta-analysis. For lifting the effect is small: a standardised mean difference of 0.20 for strength, significant for upper body but not for lower body. Expect a slightly better session rather than a different one.
Why does caffeine seem to do nothing for me?
You may genuinely be a non-responder, and there is data for that rather than just anecdote. In 101 athletes genotyped for a CYP1A2 variant, one group improved by up to 6.8%, one showed no effect at any dose, and one got measurably slower. An umbrella review also noted that not all pooled analyses gave a definite direction for the effect once prediction intervals were considered.
How long before training should I take it?
About 60 minutes is the most commonly used timing in the research, though the position stand notes the optimum depends on the source — chewing gum needs a shorter wait than a capsule because absorption differs. If you train in the evening, work backwards from bedtime rather than from the session, because the half-life will still be with you long after the last set.
Should I take caffeine before every session?
We would not. Daily use changes what a dose is doing, and much of what a habitual user feels in the morning is the restoration of their own baseline rather than an improvement on it. Saving it for the sessions that actually matter — a test, a hard interval day, a competition — keeps it working as an ergogenic aid rather than as maintenance.
Sources
- 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. View study
- 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. View study
- 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. View study
- 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. View study
- Guest N, Corey P, Vescovi J, El-Sohemy A. Caffeine, CYP1A2 Genotype, and Endurance Performance in Athletes. Med Sci Sports Exerc, 2018. View study
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
