The effects of isoflurane, halothane, and fentanyl on cerebral blood flow (CBF) and cerebral metabolic rate for oxygen (CMRO2) during anesthesia prior to carotid endarterectomy were compared using the intravenous method of 133-Xenon CBF determination. Patients, mean (+/- SE) age 68 +/- 2, received either isoflurane (N = 16), 0.75% in O2 and N2O, 50:50; halothane (N = 11), 0.5% in O2 and N2O, 50:50; or fentanyl (N = 10), 5-6 micrograms/kg bolus and then 1-2 micrograms.kg-1.h-1 infusion in addition to O2 and N2O, 40:60. Measurements were made immediately before carotid occlusion. Mean (+/- SE) CBF (ml.100 g-1.min-1) was 23.9 +/- 2.1 for isoflurane, 33.8 +/- 4.8 for halothane, and 19.3 +/- 2.4 for fentanyl. CMRO2 (ml.100 g-1.min-1) was available from 22 patients and was 1.51 +/- 0.28 for isoflurane (N = 7), 1.45 +/- 0.24 for halothane (N = 6), and 1.49 +/- 0.21 for fentanyl (N = 9). Although CBF was greater during halothane than during isoflurane or fentanyl anesthesia (p less than 0.05), there were no demonstrable differences in CMRO2 among the 3 agents. We conclude that choice of anesthetic agent for cerebrovascular surgery with comparable anesthetic regimens should not be made on the basis of "metabolic suppression." During relatively light levels of anesthesia, vasoactive properties of anesthetics are more important than cerebral metabolic depression with respect to effects on the cerebral circulation.