This study investigated the effects of two different doses of caffeine on endurance cycle time trial performance in male athletes. Using a randomised, placebo-controlled, double-blind crossover study design, sixteen well-trained and familiarised male cyclists (Mean ± s: Age = 32.6 ± 8.3 years; Body mass = 78.5 ± 6.0 kg; Height = 180.9 ± 5.5 cm VO2(peak) = 60.4 ± 4.1 ml x kg(-1) x min(-1)) completed three experimental trials, following training and dietary standardisation. Participants ingested either a placebo, or 3 or 6 mg x kg(-1) body mass of caffeine 90 min prior to completing a set amount of work equivalent to 75% of peak sustainable power output for 60 min. Exercise performance was significantly (P < 0.05) improved with both caffeine treatments as compared to placebo (4.2% with 3 mg x kg(-1) body mass and 2.9% with 6 mg x kg(-1) body mass). The difference between the two caffeine doses was not statistically significant (P = 0.24). Caffeine ingestion at either dose resulted in significantly higher heart rate values than the placebo conditions (P < 0.05), but no statistically significant treatment effects in ratings of perceived exertion (RPE) were observed (P = 0.39). A caffeine dose of 3 mg x kg(-1) body mass appears to improve cycling performance in well-trained and familiarised athletes. Doubling the dose to 6 mg x kg(-1) body mass does not confer any additional improvements in performance.