The aim of this study was to characterize neuromuscular fatigue in knee extensor muscles after a prolonged cycling exercise. During the two days preceding a 140-km race (mean +/- SD duration: 278.2 +/- 24.9 min) and 15 to 30 min after, maximal percutaneous electrical stimulations were applied to the femoral nerve of 11 trained cyclists. Electrically evoked superimposed twitches and trains of 6 stimulations were delivered during isometric maximal voluntary contraction (MVC) to determine maximal voluntary activation (% VA). Knee extensors MVC decreased with fatigue from 158.2 +/- 29.6 to 144.2 +/- 30.0 Nm (p < 0.01), but no central activation failure was detected after the race. The average rate of twitch force development increased significantly from 414 +/- 106 to 466 +/- 102 N x m x s-1 (p < 0.05) and a tendency toward higher peak twitch tension (p = 0.052) was found in the fatigued state. Short tetanus at 20 Hz and 80 Hz were also applied to 4 cyclists, but these fused and unfused tetanic forces were not significantly modified with fatigue. From these results, it can be concluded that the small but significant isometric strength loss measured less than 30 min after the end of a long distance road cycling race is not due to central fatigue. It is also suggested that a raise in peak twitch tension is not necessarily associated with enhanced neuromuscular function.