We examined the effects of time of day on a cycling time trial with and without a prolonged warm-up, among cyclists who tended towards being high in "morningness". Eight male cyclists (mean +/- s: age = 24.9 +/- 3.5 years, peak power output = 319 +/- 34 W, chronotype = 39 +/- 6 units) completed a 16.1-km time trial without a substantial warm-up at both 07:30 and 17:30 h. The time trial was also completed at both times of day after a 25-min warm-up at 60% of peak power. Power output, heart rate, intra-aural temperature and category ratings of perceived exertion (CR-10) were measured throughout the time trial. Post-test blood lactate concentration was also recorded. Warm-up generally improved time trial performance at both times of day (95% CI for improvement = 0 to 30 s), but mean cycling time was still significantly slower at 07:30 h than at 17:30 h after the warm-up (95% CI for difference = 33 to 66 s). Intra-aural temperature increased as the time trial progressed (P < 0.0005) and was significantly higher throughout the time trials at 17:30 h (P = 0.001), irrespective of whether the cyclists performed a warm-up or not. Blood lactate concentration after the time trial was lowest at 07:30 h without a warm-up (P = 0.02). No effects of time of day or warm-up were found for CR-10 or heart rate responses during the time trial. These results suggest that 16.1-km cycling performance is worse in the morning than in the afternoon, even with athletes who tend towards 'morningness', and who perform a vigorous 25-min warm-up. Diurnal variation in cycling performance is, therefore, relatively robust to some external and behavioural factors.