Temporary distal aortic perfusion and somatosensory evoked potential monitoring were attempted prospectively in 198 patients to prevent neurologic complications from occurring after operation for descending thoracic and thoracoabdominal aortic aneurysm. These 198 patients were divided into two groups. Group 1 consisted of 99 (50%) patients in whom adequate (60 mm Hg) distal bypass was achieved and combined with somatosensory evoked potential monitoring. Group 2 consisted of 99 (50%) in whom this could not be achieved. Of the latter, four patients were excluded because of operative death in one and preexisting neurologic deficits in three. The rates of early and delayed neurologic complications were 8% and 12%, respectively, in the former and 7% and 8%, respectively, in the latter. Using the logistic regression method of statistical analysis, adjusting for the difference in patient mix, we found no statistical difference in the incidence of neurologic complications in the two groups. The incidence of false negative somatosensory evoked potential response was 13% and false positive response 67%. Localization of critical spinal arteries for reattachment was not possible. Thus, the method had no significant impact upon the prevention of neurologic deficits, which varied from mild or transient to severe and which either occurred during the operation or were delayed from 12 hours to 21 days.