Computerized topographic brain mapping processes standard electroencephalographic data and displays it in a color map, thus simplifying interpretation. During a 2-year period, 65 carotid endarterectomies were performed with the use of brain mapping as the sole criterion for shunt replacement. Forty-three patients (66%) were found to have abnormal brain maps preoperatively. Ten patients (15%) developed ischemic changes after cross-clamping (all resolved after shunt placement). Postoperative brain maps were unchanged in 54 patients (83%) and improved in 7 patients (11%). A new, small focal abnormality was identified in 4 patients without shunts (6%), none of whom had a change in neurologic status. The overall major morbidity and mortality was 1.5%. Computerized brain mapping is a sensitive and readily interpretable means of monitoring cerebral perfusion during carotid surgery. We found the computerized electroencephalographic data to be a dependable criterion for selective shunting and for confirmation of shunt patency during carotid endarterectomy.