We examined whether different plasma fentanyl concentrations could maintain hemodynamic stability during coronary artery surgery. Two randomly selected groups of 10 patients were studied. Patients in group 1 received a single 75-micrograms/kg intravenous dose of fentanyl; patients in group 2 received the same dose but it was followed by an infusion of fentanyl at a rate of 0.75 micrograms/kg/min. The total dose of fentanyl in group 2 was 162 +/- 6.5 micrograms/kg (mean +/- SEM). At some point during surgery, all 10 patients in group 1 and 7 of 10 patients in group 2 had a hypertensive response. Plasma fentanyl concentrations in the two groups were not significantly different in the period 10-45 min after induction of anesthesia. At 60 min, corresponding to the time of aortic root dissection, mean plasma fentanyl concentration was statistically significantly lower in group 1 than in group 2 (13.5 +/- 1.4 ng/ml and 24 +/- 2.3 ng/ml, respectively, P less than 0.01). However, no significant difference was observed in the frequency of hypertensive response between the two groups in the period before cardiopulmonary bypass. During cardiopulmonary bypass, plasma fentanyl concentrations in group 1 were 2-3 times lower than those in group 2, and hypertension was observed in all 10 patients in group 1 but in only 2 patients in group 2 (P less than 0.05). Patients in group 1 were ventilated for 16.6 +/- 1.5 h, and those in group 2 for 19.1 +/- 1.8 h, postoperatively (P greater than 0.05). Six patients in group 1 required vasodilating drugs, whereas eight patients in group 2 required alpha-adrenergic agonists. Although the plasma fentanyl concentrations achieved were not successful in totally suppressing a hypertensive response, this study suggests that a plasma fentanyl concentration might be attainable to effect this.