Caffeine and Endurance: How Much to Take and How to Test It
A coffee before the start, a caffeine gel late in a long run: you have probably seen the "3 to 6 mg/kg" guideline without knowing where it comes from. Here is what the studies show, how to calculate your dose, and how to test it before race day.
The KILYO team
The sports coach that reads your workouts
The short answer
Yes, on average. An umbrella review of 21 meta-analyses, published in the British Journal of Sports Medicine, finds caffeine effective for aerobic endurance. In cycling time trials, a 2024 meta-analysis finds a net benefit at 4 to 6 mg/kg, not at 1 to 3 mg/kg. Calculate your dose by body weight and test it in training, well before bedtime.
Key takeaways
- An umbrella review of 21 meta-analyses (British Journal of Sports Medicine, 2020) concludes that caffeine improves aerobic endurance.
- Your dose is calculated in mg/kg: body weight multiplied by the dose, then checked on the product label.
- EFSA sets 200 mg per dose and 400 mg per day for adults, which a performance dose can exceed for a heavier athlete.
- A 400 mg dose taken 4 hours before bedtime reduced perceived sleep quality by 34%.
A coffee before a 10 km (6.2 mi) race, or a caffeine gel at the 30 km (18.6 mi) mark of a marathon: many athletes take caffeine without knowing whether it helps them, or how much to take. You often see "3 to 6 mg/kg" online, but rarely where that number comes from, or what it becomes at 55 kg or 85 kg.
Here is what recent studies say, how to calculate your dose, what Europe and the United States consider safe, and how to test it without risking your race.
What caffeine really changes during effort
The best overview is an umbrella review, which is a review of reviews. It gathers 21 published meta-analyses and concludes that caffeine is ergogenic for aerobic endurance. It also helps strength, muscular endurance, power, jumping, and speed.
The key word is "on average." Studies measure an average effect across groups. You are an individual: some athletes feel a strong effect, others feel almost nothing, and some get jitters and a racing heart. The sensible approach is to calculate a reasonable dose, then check how it affects you.
Calculating your dose in mg/kg
The math is simple: your body weight multiplied by the dose. Here is what the three most cited doses give:
| Your weight | 3 mg/kg | 4 mg/kg | 6 mg/kg |
|---|---|---|---|
| 55 kg | 165 mg | 220 mg | 330 mg |
| 65 kg | 195 mg | 260 mg | 390 mg |
| 75 kg | 225 mg | 300 mg | 450 mg |
| 85 kg | 255 mg | 340 mg | 510 mg |
Then check the label on your gel, capsule, or drink: the caffeine content is listed in mg. Don't guess it from a cup of coffee.
Low dose or medium dose
A 2024 meta-analysis of cycling time trials compared doses. At a medium dose (4 to 6 mg/kg), cyclists significantly improved their times. At a low dose (1 to 3 mg/kg), the improvement was not significant. Gains in time and average power pointed in the same direction.
Not every analysis agrees about the low dose, so don't rule it out right away. Test it.
One point about sports: these numbers come from cycling trials. For running, trail, swimming, or triathlon, coaches adapt them. Don't expect a precise percentage gain in your sport, and don't count on a specific improvement in your race time.
Safety benchmarks: Europe and the United States
In 2015, EFSA, the European Food Safety Authority, concluded that a single dose up to 200 mg (about 3 mg/kg for a 70 kg adult) poses no safety concern. It reached the same conclusion for habitual intake up to 400 mg per day in non-pregnant adults.
The US Anti-Doping Agency (USADA) also cites 400 mg per day for most adults. On both sides of the Atlantic, the daily ceiling is the same.
There is one gap to know about. Performance studies use mg/kg, while EFSA gives an absolute dose per serving. A dose of 3 to 6 mg/kg equals about 210 to 420 mg for 70 kg, which can go above EFSA's 200 mg per serving. This is only a calculation point: EFSA did not assess this specific case. At 75 kg, 6 mg/kg equals 450 mg, above the 400 mg daily limit. Stay in the lower part of the range and add up everything you drink during the day.
Not every athlete responds the same way
One possible explanation for these differences is the CYP1A2 gene, which affects how fast your body clears caffeine. A 2018 study of 101 cyclists (49 AA, 44 AC, 8 CC) measured a 10 km (6.2 mi) time trial. At 2 and 4 mg/kg, times improved only in the AA group, who are fast metabolizers. In the AC group, there was no effect. In the CC group, who are slow metabolizers, 4 mg/kg lengthened the time by 13.7% compared with placebo.
This is striking, but it is a single study with only 8 CC participants, so it is not a proven rule. You don't need a genetic test. Your best test is your own reaction in training: if at 3 mg/kg you run worse, or feel nervous or nauseous, that dose is not right for you.
When to take it, and not too late
What the studies establish is the average effect of a given dose. The exact timing of intake comes from coaching practice: you take it before the start and adjust so the peak lands when you need it, which you find out through trials. On a long run, some athletes save it for the final kilometers, when fatigue sets in.
The real trap is sleep. A 2025 crossover study of 23 healthy adult men tested 100 mg and 400 mg of caffeine, 4, 8, and 12 hours before bedtime. It concluded that 100 mg can be taken up to 4 hours before bed, while 400 mg can harm sleep.
For an evening session, it's a trade-off. 100 mg is far below a performance dose (195 mg at 65 kg for 3 mg/kg). If sleep matters for your recovery, it's better to move the session or skip caffeine that day.
What you can do before your next session
Prepare a test session like a small experiment:
- Calculate your dose in mg/kg and check the amount on the label.
- Choose a session with no stakes, not a race.
- Note the time of intake, the dose, and your last coffee.
- Change only one thing: the caffeine.
- Write one sentence about your sleep that night.
Then compare with a similar session without caffeine. KILYO reads your watch sessions and compares them: at the same pace, was your heart rate lower or higher, and did you hold your pace to the end? It can't tell what you took, so keep your notes. For a long run, also check the guide on long run fueling.
Frequently asked questions
Is caffeine allowed in competition?
Can a young athlete take it?
What about during pregnancy?
Why do I get palpitations or an upset stomach during exercise?
Sources
- Grgic J., Grgic I., Pickering C., Schoenfeld B. J., Bishop D. J., Pedisic Z. (2020). Wake up and smell the coffee: caffeine supplementation and exercise performance, an umbrella review of 21 published meta-analyses. British Journal of Sports Medicine. DOI
- Chen B., Ding L., Qin Q., Lei T.-H., Girard O., Cao Y. (2024). Effect of caffeine ingestion on time trial performance in cyclists: a systematic review and meta-analysis. Journal of the International Society of Sports Nutrition. DOI
- Guest N. et al. (2018). Caffeine, CYP1A2 Genotype, and Endurance Performance in Athletes. Medicine & Science in Sports & Exercise. Study text
- Gardiner C. L. et al. (2025). Dose and timing effects of caffeine on subsequent sleep: a randomized clinical crossover trial. SLEEP. Study text
- EFSA NDA Panel (2015). Scientific Opinion on the safety of caffeine. EFSA Journal. DOI
- USADA (2025). Substance Profile: Caffeine. United States Anti-Doping Agency. Official page
Written with the help of AI.

