Isoflurane may precipitate ischemia in patients with coronary artery disease, presumably via "coronary steal." We sought to examine whether myocardial ischemia is more common in patients with coronary artery disease receiving isoflurane (0.7%) than in a control group receiving enflurane (0.9%). Thirty-eight patients scheduled for coronary artery bypass surgery were randomly assigned one anesthetic and monitored for ischemia. All patients had ejection fractions of at least 45%. Extraneous causes of ischemia were controlled as much as possible: arterial blood pressure was maintained within 20% of baseline (primarily with phenylephrine), heart rate was maintained below 80 bpm, effects from endotracheal intubation were monitored, and measurements were made before incision. Electrocardiogram, transthoracic echocardiography, and coronary sinus lactate measurement were used to detect ischemia. Measurements were made after insertion of vascular catheters, after intubation, and after 20 min of breathing the inhaled anesthetic. During the awake period and after induction of anesthesia with fentanyl (25 micrograms/kg), there was no significant difference detected between the two groups in incidence of ischemia: 20% in the enflurane group and 22% in the isoflurane group (P = 0.38). After 20 min of receiving the inhaled anesthetic, the incidence of ischemia in the isoflurane group increased to 50%, whereas the incidence in the enflurane group was unchanged at 20% (P = 0.02). These results show that, even with strict control of hemodynamics, isoflurane is associated with more myocardial ischemia than is enflurane.