We tested the hypothesis that conduction velocity of quiescent muscle fibers decreases during sustained contraction due to the activity of the active motor units in the muscle. Ten subjects trained for the identification of a target motor unit in the abductor pollicis brevis with feedback on surface EMG signals detected with a two-dimensional array of 61-electrodes. The subjects activated the target motor unit in two 10-s long contractions, before (contraction C1) and after (C3) a 3-min long contraction (C2), all in ischemic condition. The target motor unit was not activated during C2. Eight of the ten subjects (control group) performed a second experimental session identical to the first but with a resting period of 3 min instead of the contraction C2. Exerted force and target motor unit discharge rate were not different between the two subject groups and between C1 and C3 (average over C1 and C3, mean ± SD, C2-group: 15.8 ± 10.4 % MVC and 13.1 ± 1.9 pps, respectively; control group: 15.6 ± 22.1 % MVC and 14.5 ± 1.9 pps). Muscle fiber conduction velocity of the target motor unit decreased in C3 with respect to C1 in the C2-group (3.59 ± 0.57 m/s and 3.34 ± 0.47 m/s for C1 and C3, respectively; P < 0.05) but not in the control group (3.47 ± 0.68 m/s and 3.46 ± 0.73 m/s). In the C2-group, the percent decrease in conduction velocity of the target motor unit between C1 and C3 (6.4 ± 7.1%) was not significantly different from the percent decrease in the average conduction velocity of the motor units active during C2 (9.6 ± 5.4 %). In conclusion, the contraction-induced modifications in electrophysiological membrane properties of muscle fibers are partly independent on fiber activation.