Median nerve somatosensory evoked potentials (SSEPs) were recorded in 21 healthy subjects anesthetized with halothane, isoflurane, or enflurane (with and without nitrous oxide) for abdominal or pelvic surgery. Recordings were made prior to induction, then at 0.5 MAC increments of each volatile agent with 60% N2O up to 1.5 MAC, and, finally, at 1.5 MAC without N2O. All three volatile anesthetics produced dose-related reductions in the amplitude and increases in the latency of the cortical component of the SSEP. These changes were most pronounced with enflurane and least with halothane. At 1.5 MAC of each volatile agent, cortical latency decreased and amplitude increased when nitrous oxide was discontinued. The results suggest that in neurologically intact patients, end-tidal concentrations of 1.0 MAC halothane and 0.5 MAC enflurane or isoflurane (each in 60% N2O) can be compatible with effective SSEP monitoring. Volatile anesthetic concentrations consistent with satisfactory somatosensory-evoked potential recording may be greater if N2O is not employed.