OBJECTIVE To evaluate stable xenon-enhanced CT cerebral blood flow with balloon test occlusion as a predictor of stroke risk in internal carotid artery sacrifice. METHODS Abrupt internal carotid artery occlusion was performed by surgical or endovascular means below the origin of the ophthalmic artery in 31 normotensive patients who were assessed preoperatively by a 15-minute clinical balloon test occlusion followed by an internal carotid artery-occluded xenon CT cerebral blood flow study. RESULTS One patient, who passed the clinical test occlusion but exhibited regions of cerebral blood flow less than 30 mL/100 g per minute on the occlusion xenon CT cerebral blood flow study went on to have a fatal stroke corresponding exactly to the region of reduced blood flow. Thirty patients passed both components of the preoperative stroke-risk assessment. Neuroimaging demonstrated possible flow-related infarctions, which subsequently developed in three patients. Two patients were asymptomatic, and one patient was left with a mild residual hemiparesis. CONCLUSIONS Our protocol provided a statistically significant reduction in subsequent infarction rate and infarction-related death rate when compared with a control group of normotensive abrupt internal carotid artery occlusion patients who did not undergo any preoperative stroke-risk assessment (reported in the literature). The estimated false-negative rate for our preoperative assessment protocol ranged from 3.3% to 10% depending on the assessment of the cause of the three potentially flow-related infarctions. Although life-threatening major vascular territory infarctions have been avoided, our protocol is less sensitive to changes predicting smaller, often minimally symptomatic, vascular border zone infarctions and does not predict postoperative thromboembolic strokes.