Average rectified EMG of the biceps femoris was determined. Mean dynamic flexion torque declined by 48+/-11%. Extensor maximal voluntary contraction torque, maximal torque generating capacity, voluntary activation and average rectified EMG at the four knee angles were unaffected by the hamstring fatigue protocol. Only at 50 degrees knee angle was voluntary activation significantly lower (15.7%) after fatigue ( P<0.05). In addition, average rectified EMG before fatigue was not significantly influenced by knee angle. It was concluded that a fatigued hamstring muscle did not increase the maximal voluntary contraction extension torque and knee angle did not change coactivation. Three possible mechanisms may explain the results: a potential difference in recruited fibre populations in antagonist activity compared with the fibres which were fatigued in the protocol, a smaller loss in isometric torque generating capacity of the hamstring muscle than was expected from the dynamic measurements and/or a reduction in voluntary activation.