The objective of the study is to examine the impact of accurate and inaccurate 'split-time' feedback upon a 10-mile time trial (TT) performance and to quantify power output into a practically meaningful unit of variation. Seven well-trained cyclists completed four randomised bouts of a 10-mile TT on a SRM™ cycle ergometer. TTs were performed with (1) accurate performance feedback, (2) without performance feedback, (3) and (4) false negative and false positive 'split-time' feedback showing performance 5% slower or 5% faster than actual performance. There were no significant differences in completion time, average power output, heart rate or blood lactate between the four feedback conditions. There were significantly lower (p < 0.001) averageVO 2 (ml min -1) andVE (l min -1) scores in the false positive (3,485 ± 596; 119 ± 33) and accurate (3,471 ± 513; 117 ± 22) feedback conditions compared to the false negative (3,753 ± 410; 127 ± 27) and blind (3,772 ± 378; 124 ± 21) feedback conditions. Cyclists spent a greater amount of time in a '20 watt zone' 10 W either side of average power in the negative feedback condition (fastest) than the accurate feedback (slowest) condition (39.3 vs. 32.2%, p < 0.05). There were no significant differences in the 10-mile TT performance time between accurate and inaccurate feedback conditions, despite significantly lower averageVO 2 andVE scores in the false positive and accurate feedback conditions. Additionally, cycling with a small variation in power output (10 W either side of average power) produced the fastest TT. Further psycho-physiological research should examine the mechanism(s) why lowerVO 2 andVE scores are observed when cycling in a false positive or accurate feedback condition compared to a false negative or blind feedback condition.