The effect of lidocaine on the median nerve somatosensory evoked potential (SSEP) was investigated in 14 neurologically normal patients. Lidocaine 1.5 mg.kg-1.min-1 was injected intravenously over a 5 min period immediately followed by a continuous infusion of lidocaine 60 micrograms.kg-1.min-1. The peak latencies (N1, P2, N2) and amplitudes (N1-P2, P2-N2) of the SSEP response over the sensory cortex were recorded before and after lidocaine infusion. The peak latencies in the control group and in the experimental group after lidocaine infusion of N1, P1, N2 were 19.4 +/- 1.0 msec, 19.7 +/- 1.0 msec (N1), 24.6 +/- 1.4 msec, 25.0 +/- 1.5 msec (P2), 32.5 +/- 2.5 msec, and 33.3 +/- 2.8 msec (N2), respectively. The amplitudes in the control group and in the experimental group after lidocaine infusion of N1-P2, P2-N2 were 9.0 +/- 4.3 microV, 10.3 +/- 4.7 microV (N1-P2), 7.2 +/- 3.6 microV, 8.6 +/- 3.9 microV (P2-N2), respectively. Peak latencies of all components (N1, P2, N2) increased after lidocaine infusion compared with control values. Amplitude of N1-P2 and P2-N2 increased significantly following lidocaine infusion compared with control values. The data obtained in this study suggested that the changes in peak latencies and amplitude after epidural anesthesia with lidocaine were due to the systemic effect of lidocaine absorbed intravenously from the epidural space.