Somatosensory or brain stem auditory evoked potentials (EPs) were monitored during 50 aneurysm procedures. Base line EPs were obtained shortly after the induction of general anesthesia and were recorded continuously during operation. Absolute latency, interpeak latency, and cortical EP amplitude were subsequently determined. Statistical analysis confirmed that significant changes in these parameters are routine during aneurysm operation. Arbitrary definitions of abnormal latency and amplitude changes led to an excessive false-positive/false-negative rate. From the time of dural opening to the time of closure, prolongation of central conduction time, decrease in cortical amplitude, or disappearance of the EP was predictive of postoperative sensory or motor deficit in all patients monitored, except those undergoing operation on basilar artery aneurysms. In basilar aneurysm cases, outcome could not be reliably predicted with either EP technique.