Reviews/Caffeine and Performance/Caffeine improves sprint performance
Strength Power · CAFF-PWR-02 · v2
Borderline: within 4 points of the Established/Likely boundary, so a re-grade could land either side.
Across many small, well-blinded crossover trials, caffeine tended to make athletes faster or more powerful in short, repeated bursts of maximal effort. It shortened competitive sprint times in running and swimming and raised power or total work in repeated cycling and resistance sprints, across men and women and across recreational, youth and elite athletes. The effect was not universal: one careful dose-finding trial in short cycling sprints found nothing at any dose, and another found that athletes did more total work without their peak power actually rising. Several of the positive studies also mixed caffeine with other things, such as coffee, a pre-activation warm-up, or a multi-ingredient energy drink, which makes it harder to credit caffeine alone. The panel treated the overall direction as well supported while noting these exceptions keep it from being settled beyond question.
The corpus consistently shows caffeine shortening competitive sprint time in 100-m running ([2], 0.14 s faster, P = 0.007), 60-m running ([4], 2.05% vs 0.02% in placebo) and 25/50-m swimming ([11], P = 0.032 and P = 0.033). It raises repeated-sprint power and work in Wingate and intermittent protocols ([1], [9], [3], [6], [10]) and Wingate power in resistance-trained men ([8], ES 0.33-0.44). Effects appear in independent labs and span male and female, elite, youth and recreational athletes across running, cycling and swimming. The direction of effect is favourable in nearly every included trial. The contested point: The evidence divides on whether the benefit holds for isolated short cycling sprints and on which metric actually moves: [7] found no effect on peak power (p = 0.11), mean power (p = 0.55) or time to peak (p = 0.17) at any dose 2-10 mg/kg, and [12] found total sprint number and work rose while peak and mean power did not differ. So the headline direction depends in part on the outcome measured and the exact modality tested.
All 15 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Caffeine and Performance, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
A dose-response RCT ([7], N=17) found no caffeine effect on sprint-cycling peak power (p=0.11), mean power (p=0.55) or time-to-peak (p=0.17) at any dose 2-10 mg/kg — a direct contradiction for isolated short cycling sprints.
The benefit is metric-dependent: [12] shows total sprint number and work done rose while peak and mean power did not differ, so the headline 'improves' depends on which outcome is read and does not hold uniformly.
Where it remains uncertain. A clean dose-response null for short cycling sprints ([7]) and a metric split where total work rose but peak and mean power did not ([12]) temper the breadth of the claim.
Aleatory
Crossover Ns as low as 8 ([11]) and 12-13 ([2], [3], [4], [9], [12]) make individual effect estimates noisy, including implausibly large values (g=2.27 in [1]).
Epistemic
[7] found no effect on peak power, mean power or time to peak at any dose 2-10 mg/kg, leaving the benefit for isolated short cycling sprints genuinely unresolved.
Inferential
[12] shows total sprint number and work rising while peak and mean power do not differ, so whether performance 'improved' depends on the outcome read.
The load-bearing papers from a corpus of 15, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Repeated-sprint power gains in a large triple-blind crossover.
“Caffeine increased Wmean in Wt3 (3.5%, P = 0.004, and g = 1.059) and Wt4 (3.9%, P = 0.012, and g = 1.091) compared to placebo.”
Panel auditInternal validity: high · scope of inference: Acute 3 mg/kg caffeine on repeated 30 s all-out cycle sprint power in resistance-trained men and women.Key limitations: effect confined to specific sprint numbers · single dose · cycle ergometer only
Direct competitive sprint-time improvement in trained sprinters.
“The corrected 100-m sprint time was significantly shortened by 0.14 s with caffeine supplementation compared with placebo (placebo: 11.40 ± 0.39 s, caffeine: 11.26 ± 0.33 s; P = 0.007, g = -0.33).”
Panel auditInternal validity: high · scope of inference: Acute 6 mg/kg caffeine on 100-m field sprint time in trained male sprinters.Key limitations: small sample · males only · single dose
Repeated-sprint peak power improvement in female team-sport athletes.
“Average peak power score during RST was significantly improved after LCOF (p = 0.016) and MCOF (p = 0.041) compared to PLA”
Panel auditInternal validity: medium · scope of inference: Caffeine via coffee on repeated 4 s cycle sprints in trained female team-sport athletes.Key limitations: effect only in average peak power · individual-sprint power and fatigue index unchanged · females only
40-yard dash time and RAST power metrics improved significantly in the caffeine-alone condition versus placebo.
Panel auditInternal validity: medium · scope of inference: Acute 6 mg/kg caffeine on sprint, dash and anaerobic power in recreationally active men.Key limitations: factorial design confounds · no sprint effect on standing long jump · recreational men only
4 of 15 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The directional claim is well-carried by consistent primary-RCT evidence, but a clean null and metric heterogeneity keep it short of Established.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Isolated short cycling-sprint resolution. The only clean dose-response null here ([7]) directly contradicts the broader repeated-sprint positives, leaving a specific modality unsettled. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Adequately powered, double-blind dose-response RCT of isolated maximal short cycling sprints with peak and mean power as pre-registered primary outcomes.
Open question: Does caffeine improve isolated maximal short cycling-sprint power, or is the null in [7] the true picture for that modality? Why it is still open: [7] found no effect on peak, mean or time-to-peak power at any dose, conflicting with repeated-sprint positives elsewhere in the corpus.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19