Intraoperative monitoring of central nervous system functions is useful for safe neurosurgical operations. For such a purpose, some kinds of sensory evoked potentials play significant roles, but their reliability during general anesthesia have not yet been established. The authors conducted experimental and clinical studies to reveal effects of halothane, a most popular anesthetic, on sensory evoked potentials. Eight adult cats, weighing 2.8-4.0 kg, were induced to anesthesia with thiopental and ether, and after tracheostomy and venous cannulation, they were immobilized with succinylcholine and artificially ventilated with mixture of oxygen, nitrous oxide halothane. The concentration of halothane was increased up to 4.0% by 0.5% steps. The body temperature, systemic blood pressure and carbon dioxide concentrations in expires gas were monitored continuously, and maintained within normal ranges as much as possible. The hypotension induced by halothane was easily corrected by dopamine infusion initially, but later became difficult to be controlled as the halothane concentration increased. In each concentration, short latency somatosensory evoked potentials by median nerve stimulations (SL-SEP), and brain stem auditory evoked potentials (BAEP) were recorded. Active electrode was placed on the bregma, and reference electrodes were placed on a hindlimb for SL-SEP, on the tongue for BAEP recordings. Peak amplitudes of SL-SEP were gradually decreased and finally disappeared without apparent dose dependency. Relatively the peaks with longer latencies were more affected than those with shorter latency such as P1 and N1. In BAEP, decrement of the amplitude was apparent in the peaks with longer latencies with an obvious dose dependency.(ABSTRACT TRUNCATED AT 250 WORDS)