Questions have been raised concerning the reliability of surface electrodes in achieving supramaximal stimulation during the monitoring of neuromuscular blockade; needle electrodes are considered reliable in this respect. This study compares interelectrode impedances of needle and surface electrodes during neuromuscular blockade monitoring and suggests those characteristics of the stimulation pulse that can ensure reliable supramaximal stimulation with either type of electrode. Interelectrode voltage and current for surface and needle electrodes were measured by using 1.0-ms pulses at low, medium, and high stimulation levels on 22 surgical patients during anesthesia. Data were collected immediately after electrode application, and again at 10 minutes after application. Stimulation with surface electrodes produced an initial, transient surge of current, followed by a lower steady-state value. At high stimulation levels, the peak transient current was 87% higher than the steady-state current. Needle electrodes produced a constant high-current response. At high stimulation levels the transient impedance of the surface electrode and the impedance of the needle electrode were essentially equal (0.7 k omega and 0.75 k omega, respectively). The transient impedance was significantly lower (P less than 0.001) and was associated with less interpatient variation (P less than 0.001) and less sensitivity to the duration of electrode application than was the steady-state impedance of the surface electrode. These data suggest that high-current pulses with widths of less than 0.2 ms could provide reliable supramaximal stimulation with either type of electrode.