PURPOSE: The neural mechanisms responsible for strength improvement in the early phase of strength training are unknown. One hypothesis is that strength increases because of increased neural drive to the trained muscles. Here, we used twitch interpolation to assess voluntary activation before and after a 4-wk strength training program. METHODS:: Twelve volunteers performed unilateral strength training for the right wrist abductors (three times per week). Control subjects (n = 11) practiced the same movement without resistance. We assessed voluntary activation of the trained muscles during wrist abduction and extension contractions using twitch interpolation with motor nerve and motor cortical stimulation. RESULTS:: Strength training increased wrist abduction maximal voluntary contraction (MVC) force for the trained hand by 11.0% (±8.7, P < 0.01). MVC of the untrained wrist was unchanged. There were no significant changes in wrist extension MVC force in either group. During submaximal wrist abduction, but not extension contractions, the average size of the superimposed twitches produced by cortical stimulation was significantly larger after strength training (P < 0.01). Furthermore, the direction of the twitches produced by cortical stimulation during wrist abductions and maximal wrist extension shifted toward abduction (P = 0.04). There were neither significant changes in voluntary activation measured during MVC with motor nerve or motor cortical stimulation nor changes in the amplitude of evoked EMG responses to motor cortical or motor nerve stimulation. CONCLUSIONS:: Four weeks of strength training produced a small increase in MVC that was specific to the training direction. Although maximal voluntary activation did not change with short-term strength training, the changes in direction and amplitude of cortically evoked twitches suggest that motor cortical stimulation (and presumably volition) can generate motor output more effectively to the trained muscles.