The effect of creatine loading on endurance capacity and sprint performance was investigated in elite cyclists according to a double-blind cross-over study design. Subjects (n = 12) underwent on 3 occasions and separated by 5 week wash-out periods, a 2 h 30 min standardized endurance protocol on their own race bicycle, which was mounted on an electromagnetically braked roller-system, whereupon they cycled to exhaustion at their predetermined 4 mmol lactate threshold. Immediately thereafter they performed 5 maximal 10 second sprints, separated by 2 min recovery intervals, on a Monark bicycle ergometer at 6 kg resistance on the flywheel. Before the exercise test, subjects were either creatine loaded (C: 25 g creatine monohydrate/day, 5 days) or were creatine loaded plus ingested creatine during the exercise test (CC: 5 g/h), or received placebo (P). Compared with P, C but not CC increased (p<0.05) peak and mean sprint power output by 8-9% for all 5 sprints. Endurance time to exhaustion was not affected by either C or CC. It is concluded that creatine loading improves intermittent sprint capacity at the end of endurance exercise to fatigue. This ergogenic action is counteracted by high dose creatine intake during exercise.