The aims of this investigation were to evaluate the physiological responses to laboratory based stochastic exercise and to assess the effects of stochastic versus steady-state exercise on subsequent cycling time trial (TT) performance. Six competitive cyclists (peak power output (PPO) 432 +/- 39 W (values are mean +/- SD) undertook in a random order two 150-min paced rides that were either constant load (58% of PPO) or stochastic in nature (58 +/- 12.2% of PPO). These rides were immediately followed by a 20-km TT performance on an air-braked ergometer. Mean heart rate (HR) responses throughout the 150-min paced rides and during the subsequent TT were not significantly different between trials. Yet, despite the similarities in HR, the mean time for the TT was significantly faster (26:32 +/- 1:30 vs 28:08 +/- 1:47 min, P < 0.05) and the mean power output was significantly greater (340.3 +/- 44.2 vs 302.5 +/- 42.3 W; 77.8 +/- 10.2 vs 70.0 +/- 9.8% of PPO, P < 0.05) following the steady-state ride. These results demonstrate that following 150 min of steady-state riding, subsequent 20 km TT performance was significantly improved when compared with 150 min of stochastic exercise.