Purpose: Few studies have assessed neuromuscular fatigue after self-paced locomotor exercise; moreover, none have assessed the degree of supraspinal fatigue. This study assessed central and peripheral fatigue after self-paced exercise of different durations. Methods: Thirteen well-trained male cyclists completed 4-, 20-, and 40-km simulated time trials (TTs). Pre- and immediately post-TT (<2.5 min), twitch responses from the knee extensors to electrical stimulation of the femoral nerve and transcranial magnetic stimulation of the motor cortex were recorded to assess neuromuscular and corticospinal function. Results: Time to complete 4-, 20-, and 40-km TTs was 6.0 ± 0.2, 31.8 ± 1.0, and 65.8 ± 2.2 min at average exercise intensities of 96%, 92%, and 87% of maximum oxygen uptake, respectively. Exercise resulted in significant reductions in maximum voluntary contraction, with no difference between TTs (−18%, −15%, and −16% for 4-, 20-, and 40-km TTs, respectively). Greater peripheral fatigue was evident after 4-km (40% reduction in potentiated twitch) compared with that after 20-km (31%) and 40-km TTs (29%). In contrast, longer TTs were characterized by more central fatigue, with greater reductions in voluntary activation measured by motor nerve (−11% and −10% for 20- and 40-km TTs vs –7% for 4-km TTs) and cortical stimulation (−12% and −10% for 20- and 40-km vs –6% for 4-km). Conclusions: These data demonstrate that fatigue after self-paced exercise is task dependent, with a greater degree of peripheral fatigue after shorter higher-intensity (6 min) TTs and more central fatigue after longer lower-intensity TTs (>30 min).