The present study was designed to test whether varying limb positions prior to a voluntary elbow flexion influenced the simple EMG-RTs. Eight students were examined under the conditions of six limb positions (3 × 2); three different wrist positions (45 deg. extension; WE, 45 deg. flexion; WE and neutral; N) and two different forearm positions (supination position; S and pronation position; P). Under the those six conditions, EMG-RTs and motor evoked potentials were recorded. Main results may be summarized as follows: 1) In N position, EMG-RT of S was shorter than that of P position. In P position, different EMG-RTs were observed among three different wrist positions. In S position, however, EMG-RT differences were not observed. 2) In all conditions, variabilities of the velocity and the accuracy of the movement were not found to be related to changes in EMG-RTs. 3) In N position, amplitudes of MEP in S were larger than those in P. However, variabilities of the amplitudes and the latencies of MEPs were not found to be related to changes in EMG-RTs in different wrist positions. From those results obtained, it is suggested that the change of EMG-RTs with varying forearm positions is induced dependent on two factors; one is the different number of synergic muscles taking part in an intended movement as already indicated by the "memory drum theory", another is the change of activation of the CNS which might be influenced by the change of forearm positions, and that the change of EMG-RTs with varying wrist positions is induced according to the "memory drum theory".