In active males, muscle force production and short-term (<6 s) anaerobic performance are significantly greater in the evening compared with the morning. This diurnal variation is attributed to motivational, peripheral and central factors, and possibly the higher core and muscle temperatures observed in the evening. However, little is known regarding whether diurnal variation on a treadmill also exists in team-sport specific tests of repeated sprint ability (RSA), as would be relevant to football, for example. A controlled laboratory protocol using a non-motorised treadmill has been used to investigate whether daily variation in RSA is present in highly motivated athletes. Twenty active males (mean ± SD: age, 21.0 ± 2.2 yrs; maximal oxygen uptake ([Formula: see text] max), 60.8 ± 4.8 ml kg min(-1); body mass, 77.02 ± 10.5 kg and height, 1.79 ± 0.07 m) volunteered and completed two sessions counterbalanced in order of administration (separated by >48 h): a morning (M, 07:30 h) and evening (E, 17:30 h) session. Both sessions included a 5-min active warm-up on a motorised treadmill at 10 km h(-1) followed by three task-specific warm-up sprints at 50%, 70% and 80%, respectively, on a non-motorised treadmill. During each trial, 10 × 3 s repeated sprints with 30 s recoveries were performed on the non-motorised treadmill. Rectal (Trec) and muscle temperature measurements (Tm) were taken after subjects had reclined for 30 min at the start of the protocol, and again after the active warm-up. Values of heart rate, thermal comfort (TC), rating of perceived exertion (RPE) and effort were measured throughout. Blood samples were taken at rest, after the sprints and 5-min post sprints. Data were analysed using a GLM with repeated measures. Trec and Tm values were higher at rest in the evening than the morning (0.46 °C and 0.57 °C, respectively, p < 0.05). Distance covered, peak power, average power, peak velocity and average velocity all showed significantly higher values in the evening compared with the morning (a range of 3.3-8.3%, p < 0.05), with peak power displaying a statistical trend (0.10 > p > 0.05). All subjects reported maximal values for "effort" for each sprint. There were significant positive correlations between Trec and Tm, Trec and RPE, TC and all measures of RSA performance. However, there was no correlation between fatigue index for peak power output or peak velocity and Trec. In summary, in this population of motivated subjects, time-of-day effects were seen in resting Trec and Tm values and all performance measures of RSA, in partial agreement with past research. The diurnal variation in Trec and Tm cannot fully explain time-of-day oscillations in RSA on a non-motorised treadmill. Although central temperature may provide some endogenous rhythm to human performance, the causal link seems to be due to a multiplicity of components and mechanisms.