Caffeine has one of the most robust bodies of evidence behind it of any performance-related substance, legal or otherwise. Its ergogenic effects are well-replicated across a range of sports, exercise types, and populations. It is also widely misused — overcaffeinated, poorly timed, or consumed in ways that trade short-term performance for disrupted sleep and long-term tolerance. Understanding the evidence moves this from a vague “it helps” to something more actionable.
How caffeine affects exercise performance
Caffeine’s primary mechanism is adenosine receptor antagonism. Adenosine is a neurotransmitter that accumulates during wakefulness and produces fatigue signals; caffeine blocks its receptors, reducing the perception of effort and delaying fatigue. Secondary effects include increased adrenaline release, improved neuromuscular recruitment, and enhanced fat oxidation during endurance exercise. The net result is that the same intensity of exercise feels less effortful, and effort can be sustained longer before fatigue becomes limiting.
The evidence is strongest for endurance performance. Multiple systematic reviews and meta-analyses have confirmed that caffeine produces meaningful improvements in time-to-exhaustion and time-trial performance across running, cycling, and rowing. For strength and power, the evidence is also positive but effect sizes are generally smaller. A 2024 umbrella meta-analysis confirmed ergogenic effects on both muscular strength and muscular endurance, though the authors noted that results across individual meta-analyses are still somewhat variable, partly because study designs and populations differ.
Effective doses
The most consistently studied and effective dose range is 3 to 6 milligrams per kilogram of body weight, consumed approximately 45 to 60 minutes before exercise. For a 70kg person, that is 210 to 420mg — roughly equivalent to two to four standard espresso shots depending on extraction. A 2025 network meta-analysis found moderate doses (4 to 6mg/kg) outperformed lower doses for endurance time-trial performance, though both produced meaningful improvements over placebo.
A meta-analytic review found that even very low doses — around 1 to 2mg/kg — produce ergogenic effects on resistance exercise, suggesting that minimum effective doses are lower than commonly assumed. This has practical implications: people who experience anxiety or sleep disruption from typical pre-workout caffeine doses may find a smaller amount achieves a meaningful performance benefit at lower cost to wellbeing.
Tolerance and the habituation problem
Regular caffeine consumers build tolerance relatively quickly. The adenosine receptor upregulates in response to chronic blockade, meaning the same dose produces a diminished effect over time. This is why habitual coffee drinkers often report that caffeine no longer produces noticeable alertness — they are, in effect, using it to return to baseline rather than to enhance above it.
For performance purposes, this raises a practical tension: the people most likely to use caffeine for sport are also the people most likely to have built tolerance that blunts its effects. Some research suggests a caffeine washout period — typically around a week of reduced or eliminated intake — can restore sensitivity and increase the ergogenic response. This is a contested area, and some researchers argue the tolerance picture is more complex than the binary “habituated/not habituated” framing suggests, but the principle has enough practical support to be worth considering for athletes who use caffeine strategically.
Individual variation: why caffeine hits differently
One of the more interesting and underappreciated aspects of caffeine research is the degree of individual variation in response. Two people taking the same dose will not have the same experience. A significant part of this variation is genetic. The CYP1A2 gene produces the primary enzyme responsible for metabolising caffeine in the liver. Variants of this gene determine whether someone is a fast or slow metaboliser — with fast metabolisers clearing caffeine roughly twice as quickly as slow metabolisers.
Slow metabolisers are more susceptible to anxiety, elevated blood pressure, and sleep disruption from the same dose that a fast metaboliser tolerates well. A separate gene — ADORA2A, which codes for the adenosine receptor — affects sensitivity to caffeine’s effects on sleep specifically. People with certain variants of this gene experience more pronounced sleep disruption even at moderate caffeine doses consumed well before bedtime.
These genetic differences mean that individual experimentation matters. Someone who gets anxious or cannot sleep after an afternoon espresso is not imagining it and is not weak. Their metabolic profile is genuinely different. Adjusting dose and timing accordingly is more useful than defaulting to a population-average recommendation.
Caffeine and sleep: the recovery trade-off
Caffeine has a half-life in the body of approximately five to six hours on average, though this extends considerably in slow metabolisers and shrinks in fast metabolisers. A 200mg dose consumed at 2pm will still have roughly 100mg active in your system at 7 or 8pm for an average metaboliser. For a slow metaboliser, the residual effect extends further. Given what the research shows about sleep’s role in recovery and muscle adaptation, caffeine that consistently disrupts sleep represents a meaningful long-term trade-off even if the immediate performance benefit is real.
This argues for treating afternoon and evening caffeine use carefully, particularly around training. If a session is scheduled for 6pm, the choice to caffeinate at 5pm may produce a better session at the cost of a worse night’s recovery — a net that is not obviously positive over weeks of accumulated disrupted sleep. For morning training, this trade-off largely disappears.
Form and timing
Most research uses caffeine anhydrous (capsule or powder form) rather than coffee, because coffee contains other compounds that complicate measurement of caffeine’s specific effects. That said, research comparing coffee to anhydrous caffeine at matched doses generally shows comparable ergogenic effects, suggesting the vehicle matters less than the dose. Caffeinated gum produces faster absorption than capsules and can produce a meaningful effect closer to exercise (20 to 30 minutes rather than 45 to 60), which can be practically useful when pre-exercise timing is constrained.
The bottom line on caffeine is that the evidence for its ergogenic effects is robust, the effective dose is lower than many pre-workout products suggest, the individual variation is large and partly genetic, and the interaction with sleep recovery deserves serious attention. Using it strategically — the right dose, at the right time, with an eye on its half-life and your own metabolic profile — captures most of its benefits without the common downsides.