OBJECTIVE In this study, the hemodynamics and sufentanil requirement were compared at two midazolam target plasma concentrations in patients undergoing coronary artery bypass grafting (CABG). METHODS Prospective, randomized study. METHODS University hospital, single institution. METHODS Patients undergoing CABG. METHODS Patients were randomly assigned to receive midazolam at a target plasma concentration of 150 ng/mL (group 1; n = 10) or 300 ng/mL (group 2; n = 10). Sufentanil infusion was titrated to maintain hemodynamic stability, defined as mean arterial pressure within 15% of baseline values. All patients received preoperative beta-blocking agents. Arterial blood samples of midazolam and sufentanil were analyzed by high-performance liquid chromatography and radioimmunoassay, respectively. RESULTS The mean dose of sufentanil (7.5 +/- 1.7 microgram/kg in group 1 v 7.2 +/- 2.5 micrograms/kg in group 2) did not differ. There were no significant differences in hemodynamics between the groups in the period before or after cardiopulmonary bypass (CPB). Before CPB, in two patients in each group, hypertension was controlled with sufentanil only. One patient in group 1 required a vasodilator in addition to sufentanil. No ischemic events occurred before CPB. After CPB, one patient in group 2 required a vasodilator to control hypertension. Two patients in group 2 required treatment with nitroglycerin for myocardial ischemia. Stable plasma concentrations of sufentanil and midazolam were obtained during and after CPB. The midazolam infusion was continued in both groups at a rate of 1.25 micrograms/kg/min during the first 4 postoperative hours. The time to awakening did not differ between the groups (100 +/- 58 minutes in group 1 v 173 +/- 147 minutes in group 2) nor did the plasma concentrations of midazolam (96 +/- 28 ng/mL v 108 +/- 42 ng/mL) at the time of awakening. Intraoperative awareness was not reported. CONCLUSIONS In patients undergoing CABG, good hemodynamic control with a similar incidence of hemodynamic interventions was observed at midazolam target plasma concentrations of 150 and 300 ng/mL when coadministered with sufentanil. The sufentanil requirement was identical in both groups. This study suggests that a midazolam plasma concentration of 150 ng/mL is sufficient to provide satisfactory hemodynamic control and to avoid intraoperative awareness.