The purpose of this study was to examine the reproducibility in measures of power output during a 6-s, maximal cycling sprint test. Eleven healthy, moderately-trained males (mean+/-S.D.; age=19+/-1 year; height=181.5+/-6.2 cm; mass=76.9+/-9.5 kg; peak oxygen uptake 54.9+/-6.1 mL kg(-1)min(-1)) performed a 6-s standing sprint on a front-access cycle ergometer on four separate occasions. Peak power output (PPO) was significantly higher (4.9%; P<0.05) in trial 2 compared with trial 1, whereas there were no significant differences between trials 2, 3 and 4. Similarly, the mean power output (MPO) for trial 2 was higher (5.8%; P<0.05) than in trial 1, but there were no difference across trials 2, 3 and 4. The within-subject coefficient of variation (CV) from trials 2 to 4 was 2.8 and 2.9% for PPO and MPO, respectively, while the CV calculated using data from the third and fourth trial was lower: 1.8 and 2.5% for PPO and MPO, respectively. The results of the study showed that reliable power outputs can be obtained after one familiarization session in subjects unfamiliar with maximal cycling sprint exercise. However, the inclusion of an extra familiarization session ensured more stable power outputs. Therefore, two trials should allow adequate familiarization with the maximal 6-s cycling test.