Muscle fatigue can be assessed as i) the reduction in the maximal voluntary contraction force (MVC) or ii) the inability to sustain a required level of force, i.e. endurance time (ET) of a sustained submaximal contraction until voluntary exhaustion. We tested the hypothesis that considering one or the other index may lead to different conclusions. In a first set of experiments fourteen men (28 ± 2 yrs) isometrically contracted their dominant leg knee extensor muscles at 20% of their MVC until voluntary exhaustion (ET1). At task failure, the knee extensors were electrically stimulated for 1 min with surface electrodes (40 Hz) aiming to develop the same 20% MVC target force that was not possible to develop voluntarily anymore. Potentiated doublets (100 Hz) were evoked by supramaximal stimulation of the femoral nerve before and immediately after ET1 to assess peripheral fatigue. Values are means ± S.E.M., compared by ANOVA. After task failure of ET1 (246 ± 18s), all subjects developed 20% MVC under electrical stimulation. MVC was decreased (p<0.05) by 51 ± 3% after ET1. Potentiated peak doublet was impaired after ET1 (-37 ± 4%, p<0.001) and a trend (p=0.06) towards significant correlation was observed between this reduction and the MVC decrease. In another set of experiments, thirteen men (25 ± 2 yrs) sustained 50% MVC until voluntary exhaustion with four different muscle groups (knee extensors, plantar flexors, elbow flexors and thumb adductor, dominant side) on separate occasions and MVC was measured before and immediately after exercise. ETs varied significantly (knee extensors: 77 ± 25 s; plantar flexors: 221 ± 64 s; elbow flexors: 72 ± 14 s; thumb adductor: 114 ± 27 s, p<0.05) but MVC loss (∼30-40%) was similar (p>0.05) for all muscle groups. Results of the first experiment suggest that although the large peripheral impairment seems to be associated with the reduction in maximal force generating capacity, the maximum duration of a sustained submaximal isometric knee extension performed at 20% MVC is not limited by extensor muscle force generating capacity but rather to mechanisms located above the neuromuscular junction, presumably involving descending motor drive. The results from the second experiment suggest that the MVC loss at task failure is independent of ET. Collectively, these results suggest that maximal voluntary force generating capacity and ET are two distinct indexes giving different insights into the process of muscle fatigue.