The purpose of this study was to investigate the reliability of a 90-min treadmill-based preloaded time-trial (TT (pre)) in hot (HOT) and moderate (MOD) temperatures. Nine well-trained males completed six 90-min TT (pre), three in HOT (30.5 +/- 0.1 degrees C) and three in MOD (14.4 +/- 0.1 degrees C) conditions, separated by 7 - 14 days conducted on a motorised treadmill. The TT (pre) consisted of 75 min at approximately 60 % V O (2max) followed by a self-paced 15-min time-trial (TT). Subjects ran significantly further in MOD compared to HOT during the 15-minute TT phase (p < 0.01). The CV for the 15-min TTs calculated between trials 1 and 2 were 3.3 % (MOD) and 5.8 % (HOT). This improved to 2.4 % (MOD) (p > 0.05) and significantly improved to 2.7 % (HOT) (p < 0.05) when classifying trial 1 as a familiarisation and calculating the CV between trials 2 and 3. There was no significant difference between the CV calculated between trials 2 and 3 for the 15-min TT section between conditions (p > 0.05) There were no significant differences in any physiological response between conditions or within trials of the same condition (p > 0.05). The 90-min TT (pre) is a reliable test for the assessment of endurance performance in prolonged treadmill running in familiarised males in HOT and MOD conditions.