In vivo measurements of muscle fiber conduction velocity (MFCV) have not gained wide application as compared to other electrodiagnostic study. This is partly due to a lack of clear clinical applications and partly due to the cumbersome techniques. The MFCV was determined from recordings of muscle fiber potentials with monopolar needle electrode in the fiber direction. Twenty two healthy volunteers ranging in age from 24 to 42 years served as subjects. The short head of biceps brachii of the right arm was selected for this study. Five muscle fiber potentials were recorded after distal electrical stimulation through a monopolar needle electrode at a distance of 5 cm and mean latency of each sample by averaging was calculated. Then the recording electrodes were moved proximally along the same bundles of fivers guided by the place of the twitching fibers easily found with palpation and averaged muscle fiber potentials were obtained from the motor point and 10 cm apart from the stimulation electrode by same manner. Their recording positions were adjusted until steep polypohasic potentials of and all-or-none type were recorded. The MFCV was then calculated according to the latency of the initial deflection. The mean and standard deviation of the conduction velocity were 4.25⁑0.35 m/sec (at 5 cm proximal to the stimulation electrode), 4.38⁑0.61 m/sec (at motor point) and 4.04⁑0.41 m/sec (at 10 cm proximal to the stimulation electrode), respectively. Among MFCVs, the comparison observed to be significant was that between the MFCV (over motor point) and MFCV (at 5 cm)(p<0.05), and between MFCV (at 5 cm) and MFCV (at 10 cm) (p<0.05). The MFCV recorded over motor point was significantly higher than MFCV recorded at 5 cm and 10 cm proximal to the stimulation electrode.