Ten competitive cyclists were exercised to exhaustion to test the potential of a 24-h fast for increasing endurance. One group (n = 4) was tested at an initial intensity of 86% maximum O2 uptake (VO2max) (HI) and a second group (n = 6) at 79% VO2max (MI). Both groups repeated test rides in fasted and normal-diet conditions. Time to fatigue was designated at two points: fatigue 1 occurred when pedal frequency could not be maintained at the initial percent VO2max; fatigue 2 occurred when pedal frequency could not be maintained at a workload of approximately 65% VO2max. In both HI and MI the 24-h fast had no effect on resting muscle glycogen stores but significantly increased plasma free fatty acid (FFA) levels. Despite the increased FFA availability, time to fatigue was reduced in the fasted groups. Fatigue 1 and 2 times (mean +/- SE) for HI-fasted were 42.0 +/- 6.2 and 170.0 +/- 20.4 min, respectively, compared with those of the HI-normal diet of 115.3 +/- 25.6 and 201.0 +/- 14.8 min. Fatigue 1 and 2 times for MI-fasted were 142.0 +/- 19.6 and 167.5 +/- 10.5 min compared with those of the MI-normal diet of 191.3 +/- 25.0 and 214.3 +/- 18.9 min. The cause of fatigue at fatigue 1 was not readily apparent. Fatigue 2 in all groups seemed to be related to hypoglycemia as well as muscle glycogen depletion.