Latency and amplitude characteristics of motor evoked potentials (MEPs) from abductor digiti minimi (ADM) and first dorsal interosseus (FDI) muscles were evaluated in 7 healthy volunteers via magnetic transcranial stimulation of the hemiscalp overlying contralateral motor areas. MEPs in complete relaxation and during contraction were recorded in two different experimental conditions: before and following anesthesia of median (sensory + motor) and radial (sensory) nerve fibers at wrist. This procedure induced a complete loss of skin sensation from dorsal and palmar aspects of the hand area "enveloping" the FDI muscle. On the other hand, the skin overlying the ADM muscle, as well as the strength of ulnar nerve supplied muscles were spared. This selective sensory deprivation lead to the following short-term changes: the physiological latency "jump" toward shorter values in contracted MEPs vs. relaxation was partially lost in the FDI (3.0 +/- 1.4 ms in basal condition, 1.8 +/- 1.1 ms after anesthesia, P = 0.028), while it was still clearly evident in the ADM (3.7 +/- 0.9 ms and 3.3 +/- 1.0 ms, respectively). Moreover, minor amplitude changes of MEPs during active contraction in the two muscles were detected: MEPs recorded from the FDI muscle were less potentiated during voluntary contraction than those recorded from the ADM muscle. The role of the cutaneous input in governing latency/amplitude characteristics of MEPs is discussed.