We present the results of intraoperative monitoring of motor evoked potentials from 34 patients undergoing spinal surgery under total anesthesia with intravenously administered propofol. Intraoperative recording was performed with transcranial electrical stimulation. Two groups of patients were studied: 1) a control population of 26 patients undergoing lumbar discectomy for prolapsed intervertebral disc, all of whom had normal preoperative motor conduction; and 2) a population of 8 patients undergoing neurosurgical procedures for spinal tumor (5 patients) and spinal arteriovenous malformation (3 patients), all of whom had abnormal preoperative neurological signs and abnormal preoperative motor conduction. In the first group, electromyographic responses were recorded intraoperatively either from the 2nd dorsal interosseous muscle of the hand (5 patients) or from the 1st dorsal interosseous muscle of the foot (21 patients). In the second group, responses were recorded intraoperatively either from the 1st dorsal interosseous muscle of the foot (7 patients) or from the anterior tibial muscle (1 patient). Intraoperative monitoring of motor function was successful in 88.5% of the patients in the control group. Propofol anesthesia caused a reduction in response amplitude to 7% of baseline values obtained from conscious relaxed subjects. Intraoperative monitoring was successful in 87% of the patients in the pathological group. We observed significant changes in both amplitude (greater than 50%) and/or onset latency (greater than 3 ms) from the intraoperative baseline that indicated either improvement (3 patients) or deterioration (2 patients) in motor conduction within minutes of surgical maneuvers anticipated to alter spinal cord function. Only permanent complete loss of intraoperative motor conduction (1 patient) correlated with a significant change in the postoperative neurological state.(ABSTRACT TRUNCATED AT 250 WORDS)