Fatigue during exercise is usually defined as a reversible decline in maximal muscle force or torque generating capacity (Gandevia, 2001; Allen et al., 2008). During submaximal tasks, fatigue processes during “low intensity” contractions differ from those during “high-intensity” contractions. However, whether the transition in fatigue processes occurs gradually or occurs suddenly at some identifiable “threshold” torque is unclear. This study tested the hypothesis that distinct profiles of central and peripheral fatigue would occur below and above the critical torque (CT). Following institutional ethics committee approval, nine healthy men gave written informed consent to perform seven experimental trials. Each trial involved intermittent isometric contractions of the quadriceps femoris (3 s contraction, 2 s rest; Biodex System 3 dynamometer) continued to task failure or for up to 60 min (whichever occurred the sooner), with maximal voluntary contractions (MVCs) performed at the end of each minute. Five trials were performed above CT (~35-55% MVC, denoted S1-S5 in ascending order), and two trials were performed below CT (denoted CT-20% and CT-10%). Dynamometer torque and the electromyogram (EMG) of the right vastus lateralis were sampled continuously. Peripheral and central fatigue was determined from the fall in potentiated doublet torque and voluntary activation, respectively. Above CT, contractions progressed to task failure in ~3-18 min, at which point the MVC did not differ from the target torque (Mean ± SEM: S1 target, 88.7 ± 4.3 N.m vs. MVC, 89.3 ± 8.8 N.m, P = 0.94), the average rectified EMG amplitude was not different from that measured during a MVC (S1, 76 ± 9 %MVC vs. 76 ± 8 %MVC, P = 0.97), and the potentiated doublet had fallen significantly in all trials (from 96.5 ± 5.0 to 62.5 ± 4.4 N.m, P < 0.001). Voluntary activation was also reduced in trials S1-S3, but not trials S4 and S5. During contractions below CT, contractions could be sustained for 60 min in 17 out of 18 occasions. Although the fall in MVC was significant (CT-20 from 212.4 ± 16.9 N.m to 148.3 ± 11.5 N.m, P = 0.001), a substantial reserve in MVC torque and muscle activity was evident at the end of the tasks. Potentiated doublet torque and voluntary activation were both reduced, i.e. central and peripheral fatigue developed. The rate of development of global and peripheral fatigue was 4-5 times greater during S1 than during CT-10% (e.g. ΔMVC/Δt S1 vs. CT-10%: -7.2 ± 1.4 vs. -1.5 ± 0.4 N.m.min-1 ). The results of the present study demonstrate that the rate of neuromuscular fatigue development does not scale as a simple linear function of the target torque, but is substantially accelerated above the CT.