OBJECTIVE To compare intraoperative hemodynamics profiles and recovery characteristics of propofol-alfentanil with fentanyl-midazolam anesthesia in elective coronary artery surgery. METHODS Prospective, randomized study. METHODS University hospital. METHODS Fifty patients with impaired or good left ventricular function. METHODS In group 1, (n = 25) anesthesia was induced with an infusion of propofol, 3 to 4 mg/kg/h, alfentanil, 500 micrograms, and pancuronium 0.1 mg/kg, and maintained with propofol, 3 to 6 mg/kg/h (variable rate), and alfentanil infusions, 30 micrograms/kg/h (fixed rate). Additional boluses of alfentanil, 1 mg, were administered before noxious stimuli; group 2 (n = 25) received a loading dose of fentanyl, 25 micrograms/kg, midazolam, 1.5 to 3 mg, and pancuronium, 0.1 mg/kg for induction, followed by an infusion of fentanyl, 7 micrograms/kg/h, for maintenance. Additional boluses of midazolam (1.5 to 3 mg) and fentanyl (250 micrograms) were administered before noxious stimuli. MEASUREMENTS AND MAIN RESULTS. Cardiovascular parameters at eight intraoperative time points as well as time to extubation, morphine consumption, and pain scores were recorded. Induction of anesthesia was associated in both groups with a small but significant decrease in mean arterial pressure (1: 15 mmHg (15%); 2: 8 mmHg (8%) with significant decreases in cardiac index (1: 8%; 2: 8%) and left ventricular stroke work index (1: 24%; 2: 21%). Throughout surgery, hemodynamic profiles were comparable between groups except after intubation when the MAP was significantly lower in group 1 (75 +/- 12 mmHg) than in group 2 (89 +/- 17 mmHg). Group 1 required less inotropic support. Extubation was performed faster in group 1 (7.6 h) than in group 2 (18.0 h). Morphine requirements and pain scores were comparable between groups. CONCLUSIONS Propofol-alfentanil anesthesia provides good intraoperative hemodynamics and allows early extubation after coronary artery surgery.