nutritionsciencerunningcycling

Caffeine before the start: how much you really need for a time trial

A large network meta-analysis of 48 studies shows the effective caffeine dose is more modest than commonly believed, and the performance gain is around 2%. Here is how to calculate your dose without overdoing it.

AS
Anna Severova

Caffeine is the most studied and most reliable legal supplement in endurance sport. But the question of how much to take before the start still breeds myths: some swallow a pill 'just in case,' others are convinced that more means a faster finish. A fresh network meta-analysis, published in December 2025, puts the numbers in order. And its main conclusion is likely to disappoint fans of heavy doses.

What was studied

The authors gathered 48 studies with 612 athletes (89% men) and ran a network meta-analysis. This is a method that lets you compare many intervention options at once, even if they were never tested head-to-head in a single experiment. The work included time trials in cycling (the most common format), running, rowing, and on the arm ergometer.

They compared not only doses but also the form of intake: capsules and tablets (35 studies), chewing gum (8), and mouth rinsing with a caffeine solution (5). Doses fell within the range of 2–6 mg per kilogram of body weight. Everything <2 mg/kg was excluded from the analysis as too weak, and doses >6 mg/kg simply were not tested.

Dose and timing

Here are the key numbers. The best result came from capsules in a low dose (up to 3 mg/kg): the time to cover the distance dropped by about 2.2% (SMD −0.34), and average power rose (SMD 0.38). A medium dose of 4–6 mg/kg in capsules gave about 1.8% (SMD −0.31), and medium-dose gum about 1.6% (SMD −0.30).

The most important point — there was no significant difference between the low and medium dose. This is the plateau: by adding caffeine beyond ~3 mg/kg, you buy yourself almost no extra speed, while steadily piling up the risk of side effects. The myth that 'more caffeine means a faster finish' is not backed by the numbers.

As for timing, the authors tested it as a separate factor and found no significant link with performance. In practice, for capsules people usually aim to take them about 45–60 minutes before the start — that is enough for the caffeine to be absorbed and peak in the blood.

How to apply it

Calculate your dose. The formula is simple: dose = mg/kg × your weight. For a 70 kg athlete, the 'low' dose is 70 × 3 = 210 mg. A rough guide: about two cups of strong coffee or one pharmacy caffeine tablet. Start at exactly 3 mg/kg and move up toward 6 only if you tolerate it well and feel it is not enough.

Choose the form. Capsules give a precise dose — they are easy to calculate in advance. Gum is absorbed through the lining of the mouth faster and burdens the stomach less — an option for those whose gut acts up before the start, or for taking during the race itself. Note: coffee was not studied separately in this analysis, but caffeine remains caffeine — it is just harder to measure the dose precisely in a cup.

Test it in training. Neither the dose, nor the form, nor race-day breakfast should be new. A 'washout' before the start — giving up caffeine for a few days to restore sensitivity — is a popular idea, but this particular study did not test it, so treat it as a personal experiment.

Remember sleep. Caffeine's half-life is about 4–6 hours, and longer in slow metabolizers. A dose in the second half of the day before an evening start can easily wreck your night's sleep, and lost sleep will take away more than caffeine adds. Morning races are safer in this respect.

Limitations

  • The sample is 89% male — conclusions are less reliable for women.
  • Doses >6 mg/kg and forms such as coffee and gels simply were not studied.
  • Metabolism genetics. The speed of caffeine processing depends heavily on the CYP1A2 gene: carriers of the A/C or C/C alleles are 'slow' metabolizers, in whom both the benefit and the side effects are stronger and last longer. This factor was not accounted for in the analysis, so the individual spread of response remains large.
  • Caffeine deserves extra caution with hypertension, arrhythmia, anxiety states, gastrointestinal problems, during pregnancy, and for those who sleep poorly.

The bottom line

  • The working range is 3–6 mg/kg, and it is sensible to start at the lower bound (~3 mg/kg in capsules).
  • The performance gain is modest and real — around 2% (range ~1.6–2.2%).
  • The plateau is confirmed: a medium dose is no faster than a low one, while side effects grow — there is no point chasing bigger numbers.
  • Capsules give precision, gum gives absorption speed and spares the stomach.
  • Timing did not turn out to be decisive; the guide for capsules is 45–60 minutes before the start.
  • Individual tolerance (genetics, sleep, health) matters more than a 'universal' dose — test it in training.

Source: Xue R, Huang J, Chen B, Ding L. Nutrients, 2025. DOI: 10.3390/nu17233792. https://pmc.ncbi.nlm.nih.gov/articles/PMC12694421/

Back to articles