Subjects quickly fatigue when they perform maximal voluntary contractions (MVCs). Much of the loss of force is from processes within muscle (peripheral fatigue) but some occurs because voluntary activation of the muscle declines (central fatigue). The role of central fatigue during submaximal contractions is not clear. This study investigated whether central fatigue developed during prolonged low-force voluntary contractions. Subjects (n=9) held isometric elbow flexions of 15% MVC for 43 min. Voluntary activation was measured during brief MVCs every 3 min. During each MVC, transcranial magnetic stimulation (TMS) was followed by stimulation of either brachial plexus or the motor nerve of biceps brachii. After nerve stimulation, a resting twitch was also evoked before subjects resumed the 15% MVC. Perceived effort, elbow flexion torque and surface EMG from biceps, brachioradialis and triceps were recorded. TMS was also given during the sustained 15% MVC. During the sustained contraction, perceived effort rose from approximately 2 to approximately 8 (out of 10) while ongoing biceps EMG increased from 6.9+/-2.1% to 20.0+/-7.8% of initial maximum. Torque in the brief MVCs and the resting twitch fell to 58.6+/-14.5 and 58.2+/-13.2% of control values, respectively. EMG in the MVCs also fell to 62.2+/-15.3% of initial maximum, and twitches evoked by nerve stimulation and TMS grew progressively. Voluntary activation calculated from these twitches fell from approximately 98% to 71.9+/-38.9 and 76.9+/-18.3%, respectively. The silent period following TMS lengthened both in the brief MVCs (by approximately 40 ms) and in the sustained target contraction (by approximately 18 ms). After the end of the sustained contraction, the silent period recovered immediately, voluntary activation and voluntary EMG recovered over several minutes while MVC torque only returned to approximately 85% baseline. The resting twitch showed no recovery. Thus, as well as fatigue in the muscle, the prolonged low-force contraction produced progressive central fatigue, and some of this impairment of the subjects' ability to drive the muscle maximally was due to suboptimal output from the motor cortex. Although caused by a low-force contraction, both the peripheral and central fatigue impaired the production of maximal voluntary force. While central fatigue can only be demonstrated during MVCs, it may have contributed to the disproportionate increase in perceived effort reported during the prolonged low-force contraction.