Carotid endarterectomy requires temporary surgical occlusion of the involved carotid artery. During occlusion, the minimally acceptable (critical) internal carotid artery stump pressure is reported to be 50 torr, whereas for regional cerebral blood flow (rCBF), a critical range is reported to be 18-24 ml/100 g/min. During 90 carotid endarterectomies, rCBF and stump pressure were measured and the EEG continuously monitored. A positive correlation between rCBF and stump pressure (i.e., when both were either above or below their respective critical values) was observed in only 58 per cent of the cases. In 28 per cent stump pressures of less than 50 torr were observed despite rCBF's above 24 ml/100 g/min and normal EEG's. In 8 per cent stump pressures were more than 50 torr but rCBF's were less than 18 ml/100 g/min and EEG changes of ischemia were commonly observed. In the remaining 6 per cent rCBF's were marginal (18-24 ml/100 g/min) while stump pressures were more than 50 torr and EEG changes were not observed. The relationship between stump pressure and rCBF was influenced by the anesthetic. In the absence of transient ischemia during occlusion (that is, rCBF greater than 18 ml/100 g/min), halothane and enflurane anesthesia were associated with significantly higher rCBF's and lower stump pressures than was neuroleptanesthesia. Pre-occlusion and post-occlusion rCBF measurements also demonstrated cerebral vasodilation by halothane and enflurane (halothane greater than enflurane) and vasoconstriction by neuroleptanesthesia. It is concluded that stump pressure is an unreliable index of CBF during carotid occlusion and that its relationship to CBF is considerably influenced by the anesthetic.