The main purpose of this study was to examine whether facilitation of human motor unit recruitment by repeated voluntary contractions is mediated, in part, by time and activity-dependent increases in the intrinsic excitability of the parent motoneuron. To do this, pairs of tibialis anterior or soleus motor units were recorded during slowly increasing and then decreasing voluntary contractions. The firing rate of the lower-threshold motor unit of the pair (control unit) was used as a measure of effective synaptic excitation (i.e., drive) to the motoneurons. This rate was used to estimate the recruitment threshold of the higher-threshold unit of the pair (test unit). The test unit was repeatedly recruited and de-recruited in a series of contractions, and the interval between the de-recruitment and re-recruitment of the test unit (interactivation interval) was systematically varied between 0.6 and 60 s. An increase in intrinsic excitability of a unit was considered to have occurred if the level of estimated synaptic input (as measured by the firing rate of the control motor unit) needed to recruit a unit was reduced. At short interactivation intervals (1-2 s), the control unit firing frequency was significantly lower when the test unit was recruited on the second contraction, compared with the first (by 3.9 Hz or a 64% reduction). This suggested that the intrinsic excitability of the test motor unit had increased during the second contraction because it could be recruited at a much lower level of estimated synaptic drive. Longer interaction intervals (2-6 s) produced less recruitment facilitation. At even longer interactivation intervals (>6 s) there was no significant facilitation (time constant of effect was 4.8 s). In some motor units, the effect of this short-term facilitation appeared to be so pronounced that it resulted in reversing the order of de-recruitment with the other initially lower-threshold motor units. Such reversals were occasionally observed for orderly re-recruitment. The time course and behavior of the observed short-term facilitation of motor unit discharge was qualitatively similar to the warm-up phenomenon of plateau potentials seen in motoneurons of reduced preparations (e.g., 4-6 s). The possibility of warm-up contributing to the time and activity-dependent facilitation of human motor unit recruitment is discussed.