Endurance performance is a common criterion used to evaluate training or dietary interventions. However, to accurately appraise the effects of an intervention, the endurance performance measure must be reliable. The purpose of the investigation was to establish the reliability of a 1-h endurance performance test. Twenty trained female subjects (peak VO2 = 47.4 +/- 7.2 ml.kg-1.min-1) completed two trials in which they had to generate the highest power output possible throughout 60 min of cycling. Heart rates (HR) and ratings of perceived exertion (RPE) were also recorded during these two trials. All tests were conducted on a wind-braked cycle ergometer set up to closely resemble the subject's own cycle. The trials were separated by 1 wk, conducted on the same day of the week, and completed at a similar time of the day. The average power outputs (+/-SD) for the two trials were 180.0 (+/-18.1) W and 180.0 (+/-20.6) W. The results revealed that average absolute power output, HR, and RPE were not significantly different between trials. The intraclass correlation coefficient (one way ANOVA) for average absolute power output was 0.97, the coefficient of variation was 2.7%, and the SEM was 3.4 W. These results suggest that under controlled conditions average absolute power output during a 1-h endurance test is a reliable measure for trained female cyclists.