Reviews/Caffeine and Performance/Low doses improve endurance
Dosing · CAFF-DOS-02 · v2
A handful of small controlled studies suggests that a low dose of caffeine, around 3 mg per kilogram of body weight, can give a short-term lift to endurance exercise. One cycling study found people did more work in a single session after caffeine [1], and a treadmill study found athletes could keep going longer before exhaustion [7]. But the evidence is thin and uneven: the trials that actually tested aerobic endurance were very small, the single largest study measured a different kind of endurance (how long muscles can keep lifting, not sustained aerobic effort) [11], and one proper endurance running test barely moved at all [3]. The cycling benefit also disappeared after four weeks of daily use as tolerance developed [1]. So the direction points toward a benefit, but the base is narrow and comes with important caveats, which is why the panel stopped short of a confident conclusion.
Two small crossover trials give direct aerobic-endurance signal at a low dose: in [1] an acute 3 mg/kg dose raised cycling external work (383.3 ± 75 kJ vs. 344.9 ± 80.3 kJ; ES = 0.49; P = 0.001), and in [7] both 3 and 6 mg/kg significantly increased treadmill time-to-exhaustion (P < 0.05). A large triple-blind trial [11] (n=94) reports that acute 3 mg/kg caffeine enhances muscular endurance to failure at 65% 1RM, though this is a resistance construct rather than aerobic endurance. Within the one soccer trial carrying an intermittent-endurance test [3], Yo-Yo IR1 distance was 2440 (531) m vs 2308 (540) m at P = 15%, the weakest effect in that study. The surrounding corpus largely tests sprint, jump, agility and strength rather than endurance. The contested point: The evidence divides on whether the low-dose benefit generalises to endurance: the directional support rests on two small aerobic trials [1], [7] plus a muscular-endurance proxy [11], while the single dedicated endurance running test [3] was essentially null (P = 15%). It also divides on durability, since the cycling benefit in [1] vanished after four weeks of low-dose intake.
All 13 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.
Only two papers ([1] n=18, [7] n=12) test aerobic endurance at a low dose; both are small crossover trials vulnerable to a few responders, a thin base for a general directional claim.
The largest, best-powered support [11] n=94 measures muscular endurance (reps/velocity to failure), not aerobic endurance; crediting it to the claim conflates two distinct constructs and inflates apparent evidence.
Where it remains uncertain. The best-powered support [11] n=94 measures muscular rather than aerobic endurance, the one dedicated endurance test [3] moved only at P = 15%, and the [1] benefit disappeared after four weeks of intake.
Aleatory
The aerobic trials are small crossovers ([1] n=18, [7] n=12), so a few responders can swing the group effect.
Epistemic
The acute cycling benefit in [1] was lost after 4 weeks of low-dose intake (postcaf ES = 0.31; P = 0.025; placebo retained), so whether 'improve' holds beyond acute, non-habituated use is unestablished.
Inferential
Reading muscular endurance to failure in [11] as evidence for aerobic endurance substitutes a distinct construct for the one the claim names.
The load-bearing papers from a corpus of 13, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Acute 3 mg/kg low-dose caffeine improved external work in a cycling endurance task (60 min at 60% V̇O2peak + 30 min performance task).
“This performance benefit was no longer apparent after 4 weeks of caffeine supplementation (precaf: 383.3 ± 75.0 kJ vs. postcaf: 358.0 ± 89.8 kJ; ES = 0.31; P = 0.025)”
Panel auditInternal validity: medium · scope of inference: Acute low-dose caffeine in low-habitual consumers performing submaximal cycling endurance, and loss of that benefit after 4 weeks of chronic low-dose intake.Key limitations: small sample · low-habitual consumers only · acute benefit lost after chronic use
Counter-evidence: the corpus's intermittent-endurance test showed the lowest probability of improvement, unlike sprint/agility outcomes.
“Yo-Yo IR1 distance was 2440 (531) m vs 2308 (540) m ( P = 15%)”
Panel auditInternal validity: medium · scope of inference: Low-dose caffeine on sprint, agility, jump and intermittent-endurance tasks in elite adolescent male soccer players.Key limitations: single-blind · endurance test showed only 15% probability of improvement · small adolescent sample
Low-dose 3 mg/kg (C3) significantly increased an aerobic time-to-exhaustion endurance outcome.
“C3 and C6 significantly increased time-to-exhaustion following treadmill testing ( P < 0.05)”
Panel auditInternal validity: medium · scope of inference: Acute low-to-moderate caffeine on time-to-exhaustion, VO2max and anaerobic power in male kickboxing athletes.Key limitations: blinding not stated · very small sample · no VO2max effect
3 of 13 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.
Direction favours the claim from direct primary trials, but small samples, acute-only benefit with tolerance, a within-endurance null, and reliance on a muscular-endurance proxy keep confidence below the Likely band.
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.
Adequately powered aerobic-endurance trial at low dose. Direct aerobic-endurance support rests on only two small crossovers ([1] n=18, [7] n=12), too thin for a general directional claim. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A larger, pre-registered randomised crossover of 3 mg/kg vs placebo on a validated sustained aerobic-endurance test (time-trial or time-to-exhaustion) powered for a modest effect.
Open question: Does a low dose of caffeine improve sustained aerobic endurance in a trial large enough to rule out a few-responder effect? Why it is still open: The only direct aerobic trials are small crossovers ([1] n=18, [7] n=12).
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19