This study determined the contribution of supraspinal fatigue and contractile properties to the age difference in neuromuscular fatigue during and recovery from a low-intensity sustained contraction. Cortical stimulation was used to evoke measures of voluntary activation and muscle relaxation during and after a contraction sustained at 20% of maximal voluntary contraction (MVC) until task failure with elbow flexor muscles in 14 young adults (20.9 ± 3.6 yr, 7 men) and 14 old adults (71.6 ± 5.4 yr, 7 men). Old adults exhibited a longer time to task failure than the young adults (23.8 ± 9.0 vs. 11.5 ± 3.9 min, respectively, P < 0.001). The time to failure was associated with initial peak rates of relaxation of muscle fibers and pressor response (P < 0.05). Increments in torque (superimposed twitch; SIT) generated by transcranial magnetic stimulation (TMS) during brief MVCs, increased during the fatiguing contraction (P < 0.001) and then decreased during recovery (P = 0.02). The increase in the SIT was greater for the old adults than the young adults during the fatiguing contraction and recovery (P < 0.05). Recovery of MVC torque was less for old than young adults at 10 min post-fatiguing contraction (75.1 ± 8.7 vs. 83.6 ± 7.8% of control MVC, respectively, P = 0.01) and was associated with the recovery of the SIT (r = −0.59, r2 = 0.35, P < 0.001). Motor evoked potential (MEP) amplitude and the silent period elicited during the fatiguing contraction increased less for old adults than young adults (P < 0.05). The greater fatigue resistance with age during a low-intensity sustained contraction was attributable to mechanisms located within the muscle. Recovery of maximal strength after the low-intensity fatiguing contraction however, was impeded more for old adults than young because of greater supraspinal fatigue. Recovery of strength could be an important variable to consider in exercise prescription of old populations.