Caffeine sharpened sprints, not jumps or strength

Caffeine is one of the most used ergogenic aids in sport, but most of its evidence base comes from cycling, running or mixed team-sport populations. Rugby is different: it demands repeated high-intensity sprinting, technical skill execution and physical collision, all performed under accumulating fatigue, so findings from other sports may not transfer. Coaches and players also need to know whether any performance edge comes at a cost to sleep and recovery. This review set out to pool the randomised evidence specifically in rugby players across physical performance, sport-specific skill, perceptual responses and physiological markers.

The authors searched PubMed, Web of Science, the Cochrane Library, Scopus and Embase from inception to May 2026 for randomised placebo-controlled studies of acute caffeine supplementation in rugby players; observational studies on sleep and recovery were summarised narratively. Effects were expressed as Hedges' g with 95% confidence intervals, using three-level random-effects models to handle dependent effect sizes. Methodological quality and risk of bias were assessed with the PEDro scale and RoB 2, observational studies with the JBI checklist, and overall certainty with GRADE. Fifteen studies were included, with 13 randomised trials contributing meta-analytic data.

Acute caffeine was associated with small improvements in sprint performance (Hedges' g = -0.41 across 11 effects) and lower ratings of perceived exertion measured during or immediately after rugby-specific or resistance-exercise protocols (g = -0.40 across 9 effects). Passing accuracy showed a large positive effect (g = 1.65), but with high heterogeneity and signs of small-study effects, so it is the least trustworthy result. There were no significant effects on jumping performance, strength performance, metabolic markers or hormonal markers. Subgroup analyses found no moderation by rugby code, dose, form of administration or competitive level, and observational evidence pointed to possible sleep and recovery concerns.

This confirms the familiar pattern that caffeine's clearest effects sit in sprint-type output and in how hard exercise feels, rather than in maximal strength or jump power, and it extends that pattern to a collision sport. The striking passing-accuracy result is the novel claim, but its heterogeneity and small-study signal mean it needs replication before anyone treats caffeine as a skill enhancer. Certainty of evidence ranged from low to very low throughout. The failure of dose, form and competitive level to moderate outcomes leaves the practical question of who benefits, and how much, unresolved.

For team-sport athletes, this is a reason to think caffeine may make repeated sprinting slightly faster and hard sessions feel slightly easier, while doing little for jumping or maximal strength. The eye-catching passing-accuracy finding should be read with real caution given how unstable it was across studies. The review also notes observational evidence hinting at knock-on effects for sleep and recovery, which matters if caffeine is used late in the day around fixtures. Overall the honest summary is a small, real perceptual and sprint effect on low-certainty evidence.