Sports nutrition is an industry built substantially on weak evidence and confident marketing. Most supplements marketed for performance have thin support, small studies, or effects that vanish under proper controls. Caffeine is the conspicuous exception, and it is worth understanding exactly how strong its evidence base is, because it is stronger than almost anything else you can legally buy.
It is also worth knowing that caffeine was on the World Anti-Doping Agency's prohibited list until 2004, when it was moved to the monitoring programme. It was banned precisely because it works, and it was unbanned largely because a substance present in ordinary coffee is impractical to police. Very few legal supplements have ever been considered effective enough to prohibit.
The umbrella review
The best single summary is by Grgic and colleagues in the British Journal of Sports Medicine in 2020[1], an umbrella review pooling 21 published meta-analyses of caffeine and exercise performance. An umbrella review sits at the top of the evidence hierarchy: rather than pooling individual trials, it pools the pooled analyses, which surfaces whether a result is consistent across independent teams asking slightly different questions.
Their conclusion was that caffeine ingestion improves exercise performance across a broad range of tasks, with ergogenic effects on muscle endurance, muscle strength, anaerobic power and aerobic endurance. Those effects were substantiated by moderate quality of evidence drawn from moderate-to-high quality systematic reviews. In sports nutrition, that grading is close to as good as it gets.
One qualification matters: the magnitude of the effect is generally greater for aerobic exercise than for anaerobic exercise. Caffeine helps a runner or cyclist more than it helps a sprinter or a single maximal lift, though it helps both.
The position stand
The practical companion is the International Society of Sports Nutrition position stand on caffeine and exercise performance[2], published in 2021 with Grgic among its authors. A position stand translates the literature into usable recommendations, and this one is the standard reference in the field.
The commonly supported dosing range is roughly 3 to 6 mg per kilogram of body weight, taken around 60 minutes before exercise. For a 70 kg person that is about 210 to 420 mg. Higher doses, above 9 mg per kilogram, generally do not improve performance further and increase side effects, so more is not better here either. The effect appears across trained and untrained individuals and is not restricted to elite athletes.
Does coffee work, or do you need a pill?
This is the question most people actually want answered, and the research supports coffee. Studies comparing caffeine delivered as coffee against caffeine delivered in anhydrous form have generally found comparable ergogenic effects when the caffeine dose is matched. The position stand reflects this, treating coffee as a legitimate delivery vehicle.
The practical complication is dose control. An espresso shot contains roughly 63 mg of caffeine and an 8-ounce brewed coffee roughly 95 mg, but real cups vary enormously with bean, roast, grind, brew method and size. A 70 kg person aiming at 4 mg per kilogram wants about 280 mg, which is roughly three cups of brewed coffee, and three cups is a lot of liquid to take on an hour before training. This is the main reason competitive athletes tend to use capsules: not because coffee is less effective, but because it is harder to dose precisely and comes with a volume of fluid.
It is also worth knowing that the diterpenes discussed in our article on unfiltered coffee and cholesterol are irrelevant to performance, so brewing method does not matter here except insofar as it changes caffeine content. Espresso-based drinks are the efficient option if volume is a concern.
Why it works
The dominant mechanism is central rather than muscular. Caffeine blocks adenosine receptors in the brain, which reduces the perception of effort and of pain during exercise. The same work feels easier, so people sustain a higher output for longer. There is also evidence for effects on motor unit recruitment and on calcium handling within muscle, but the perception-of-effort pathway is the best supported and the most consequential.
This has an implication people find counterintuitive: caffeine is not adding energy. It is changing how hard the effort feels, which lets you access output you already had. That framing also clarifies why the benefit is larger in endurance events, where perceived effort accumulates over a long duration, than in a single maximal lift.
The complications worth knowing
Habituation is the first question everyone asks, and the evidence is genuinely mixed. Some studies find the ergogenic effect blunted in habitual heavy consumers, others find it preserved. The common recommendation to abstain for several days before a competition to restore sensitivity has weaker support than its popularity implies, and abstaining brings its own problem, since Juliano and Griffiths[3] documented that withdrawal symptoms include fatigue, reduced alertness and headache, none of which help on race day.
Getting the dose right in practice
The 3 to 6 mg per kilogram range needs translating, because most people have no sense of what that means in cups. A 60 kg person is looking at 180 to 360 mg; a 75 kg person at 225 to 450 mg; a 90 kg person at 270 to 540 mg.
Against an 8-ounce brewed coffee at roughly 95 mg and a single espresso shot at roughly 63 mg, the low end of the range for a 60 kg person is about two cups of brewed coffee, and the high end for a 90 kg person is well beyond what anyone would drink in one sitting an hour before training. This is the practical reason serious athletes use capsules: the upper half of the effective range is difficult to reach through coffee without also consuming an uncomfortable volume of liquid.
The sensible approach for anyone not competing is to start at the low end, around 3 mg per kilogram, which is achievable with coffee, and see whether it helps before considering more. The dose-response curve flattens well before the top of the range, so the extra is mostly extra side effects.
What caffeine does not do
It is worth being clear about the boundaries, because the supplement industry is not. Caffeine does not build muscle, does not increase your maximum strength in any structural sense, and does not substitute for training, sleep or food. What it does is reduce the perceived cost of effort, which lets you express capacity you already have.
That distinction matters practically. If you are undertrained, underslept or underfed, caffeine will make a session feel more manageable without making it more productive, and it can mask fatigue signals that exist for a reason. Used that way it borrows against recovery rather than adding to performance.
Genetics is the second. Nehlig's review[4] documents substantial interindividual variation in caffeine response driven by CYP1A2 and adenosine receptor variants, and some studies have reported that fast metabolisers gain more ergogenic benefit than slow ones, who may even be impaired at higher doses. The evidence here is not yet strong enough to justify genotyping before deciding how to dose, but it is a real reason why individual results vary so much. Our article on caffeine genetics covers this.
Sleep is the third and most underrated. Training performance across a week depends heavily on recovery, and Drake and colleagues showed that caffeine 6 hours before bed cut objectively measured sleep by over an hour[5]. An evening session fuelled by a pre-workout dose can easily cost more in recovery than it gained in that session's output. If you train late, this trade-off is the most important thing on this page for you.
There is a fourth issue specific to coffee as a delivery vehicle, and it is gastric rather than pharmacological. Several hundred millilitres of coffee an hour before exercise is a meaningful volume of acidic liquid, and for people prone to reflux or gastrointestinal distress during exercise, that alone can outweigh any ergogenic gain. This is another reason capsules or espresso-based drinks are preferred by people training seriously, and it has nothing to do with the caffeine itself.
It is also worth flagging that caffeine's effect on thermoregulation and fluid balance is frequently overstated in the other direction. Caffeine at these doses does not meaningfully dehydrate a habituated user, and the older advice to offset coffee with extra water is not supported by current evidence.
The safety boundary
Performance doses sit close to general daily limits. The FDA[6] identifies 400 mg a day as the level not generally associated with negative effects in healthy adults, and EFSA reached the same 400 mg daily figure, with single doses up to 200 mg not raising safety concerns[7]. A 6 mg per kilogram dose for an 80 kg athlete is 480 mg in one sitting, which exceeds both. That does not make it acutely dangerous for a healthy adult, but it is well outside routine-consumption guidance and should not be a daily habit. Anyone pregnant, with a cardiac condition, or with an anxiety disorder should treat performance dosing as off-limits without medical advice.
The summary
Caffeine improves muscle endurance, strength, anaerobic power and aerobic endurance, established across 21 meta-analyses, with the largest effects in aerobic exercise. The supported dose is about 3 to 6 mg per kilogram roughly an hour beforehand, coffee works as a delivery vehicle when the dose is matched, and the mechanism is mainly a reduction in perceived effort rather than added energy.
Where this leaves an ordinary coffee drinker
Most people reading this are not competing at anything, so it is worth extracting what actually applies. If you exercise in the morning, a cup of coffee 45 to 60 minutes beforehand is a well-evidenced, free, low-risk way to make the session feel easier, and there is no need to weigh anything or calculate milligrams per kilogram.
The umbrella review's findings apply to trained and untrained people alike, so the effect is available to ordinary exercisers, not just athletes. What differs is how much it matters: shaving perceived effort off a race is worth more than shaving it off a Tuesday jog, though the second is a perfectly good reason to have a coffee first.
If you exercise in the evening, the calculation flips, and the sleep evidence in our article on caffeine and sleep should dominate. A 7pm pre-workout coffee is inside the six-hour window for almost anyone, and the recovery cost is likely to exceed the session's gain.
If you already drink coffee all day, you are unlikely to get a dramatic acute effect from another cup before training, because whatever tolerance you have developed is already in place. The people who notice caffeine's ergogenic effect most vividly are those who use it rarely.
And if caffeine makes you anxious or shaky, that will not improve under exercise, where heart rate and adrenaline are already elevated. Perceived exertion falling is of limited use if you feel unwell for the whole session.
The caveats are habituation of uncertain size, real genetic variation in response, and a serious sleep trade-off for anyone training in the evening. For the everyday version, our beginner's guide covers how much caffeine is actually in different drinks.
Weighed against the rest of sports nutrition, though, caffeine remains the one supplement where the question is how best to use it rather than whether it does anything at all.
Timing, and why 60 minutes
The standard recommendation of roughly 60 minutes before exercise comes from caffeine's absorption profile: plasma concentrations generally peak somewhere between 30 and 90 minutes after ingestion, with considerable variation between individuals and between delivery forms.
That variation is worth accounting for rather than treating the hour as precise. Caffeine taken with a substantial meal absorbs more slowly; taken on an empty stomach it peaks sooner. Chewing gum formulations absorb through the mouth and act faster, which is why they are used in events where timing is tight. And for a long endurance event, peak concentration matters less than sustained levels, so the timing question softens considerably.
For an ordinary person having a coffee before a workout, anywhere from 30 to 60 minutes ahead is fine, and the difference between the two is smaller than the difference between having it and not.