An analysis was undertaken of 458 consecutive carotid endarterectomies performed over 6 years with the patient under general anesthesia and with electroencephalographic monitoring. Seventy patients (15%) had electroencephalographic changes suggestive of ischemia with carotid clamping and had shunts placed. Ischemic encephalographic changes occurred in 26% of patients with an occluded contralateral carotid artery, 21% of patients with a prior stroke history, and 12% of patients with no stroke history and a patent contralateral carotid artery. Nineteen strokes (4.1%), nine transient deficits (2.0%), and one death (0.2%) occurred in the 458 endarterectomies in this experience. Ten of the 19 strokes and five of nine transient deficits were immediately apparent when patients awoke from anesthesia. Five of 10 patients with immediate strokes and all five patients with immediate transient deficits had no ischemic electroencephalographic changes during the procedure. Two other patients with immediate strokes initially had ischemic electroencephalographic changes after carotid clamping that reversed with increased blood pressure or shunting. Therefore 7 of 10 patients with immediate strokes and all 5 patients with immediate transient deficits had electroencephalographs unchanged from baseline at completion of the procedure, and thus deficits not manifest by operative electroencephalographic changes developed. Our data do not support the tenet that electroencephalographic monitoring will always predict neurologic deficits accompanying carotid endarterectomy.