Global and regional myocardial blood flow and metabolism were examined in 20 patients with coronary artery disease before surgical stimulation. Half were anesthetized with halothane (0.8%) and half with isoflurane (1.2%). Coronary perfusion pressure decreased similarly in both groups. During halothane anesthesia coronary sinus blood flow, an index of global perfusion, decreased from an awake value of 129 +/- 7 to 97 +/- 7 ml/min (P less than 0.05), and great cardiac vein blood flow, an index of regional perfusion, decreased from 60 +/- 8 to 44 +/- 5 ml/min (P less than 0.05). In contrast, during isoflurane anesthesia global coronary blood flow increased from 131 +/- 13 to 153 +/- 16 ml/min (P less than 0.05), while regional blood flow decreased from 68 +/- 7 to 56 +/- 6 ml/min (P less than 0.05). Thus, the ratio of great cardiac vein blood flow to coronary sinus blood flow was unchanged during halothane anesthesia but decreased significantly during isoflurane. Neither global nor regional coronary vascular resistance was altered by halothane, whereas isoflurane decreased global coronary vascular resistance without affecting regional coronary vascular resistance. All patients receiving halothane had net myocardial lactate extraction. In the isoflurane group, four patients showed global lactate production and three regional lactate production. All patients demonstrating lactate production also developed electrocardiographic evidence of myocardial ischemia, which was not present before induction. The authors conclude that halothane is a preferable anesthetic to isoflurane in patients with coronary artery disease because the latter has the propensity to induce maldistribution of the coronary circulation and myocardial ischemia.