This study monitored the effects of a short-term elbow flexor training program on surface electromyographic (SEMG) spike activity. The experimental paradigm consisted of three test sessions separated by 2-week intervals. At the beginning of each session, participants (N=13) performed five maximal effort isometric contractions of the elbow flexors to serve as baseline. After 5 min of rest, the participants then engaged in a 30-trial isometric fatigue protocol during which maximal elbow flexion torque was measured with a load-cell, and the maximal rate of change in the torque (dtau/dt(max)) was obtained from the differentiated torque-time curve. Bipolar electrodes were used to monitor the SEMG spike activity of the biceps brachii. Mean spike amplitude (MSA) and mean spike frequency (MSF) were calculated for the torque development and constant-torque phases of the isometric contraction, termed Segment 1 and Segment 2, respectively. Mean power frequency (MPF) was also calculated for Segment 2. The five baseline contractions of the second and third sessions were compared with those of the first session and analyzed for training-related changes. Training increased dtau/dt(max) but failed to change maximal elbow flexion torque or MSA. However, there was an increase in the MSF during the torque development phase of the contraction (Segment 1). Both MSA and MSF were greatest during the constant-torque phase of the isometric contraction (Segment 2). There was a strong linear correlation (r=0.90, P<0.05) between MSF and MPF during (Segment 2). We hypothesize that the increase in dtau/dt(max) is due to enhanced motor-unit rate-coding. The demonstrated correlation between MSF and MPF measures will allow investigators to use spike analysis to examine the frequency content of the SEMG signal under non-stationary conditions.