We conducted a prospective, randomized study to compare quality of sedation, hemodynamic stability, and oxygen consumption of three different drugs for continuous IV sedation in the immediate postoperative period in patients scheduled for aortic surgery (propofol [n = 12], sufentanil [n = 12], or midazolam [n = 12]). After arrival in the recovery room, patients were randomized into one of the following groups: Group P (continuous infusion of propofol 2mg.kg(-1).h(-1)), Group S (continuous infusion of sufentanil 0.25 mu g.kg(-1).h(-1), with bolus doses of midazolam 2 mg to maintain sedation at 3-4 on the Ramsay scale), and Group M (continuous infusion of midazolam 0.07-0.15 mu g.kg(-1).h(-1) to maintain sedation at 3-4 on the Ramsay scale). The three drugs were associated with similar hemodynamic stability, incidence of myocardial ischemia, and comparable kinetics and mean values for Vo(2), but a significant higher number of peaks for Vo(2) in Group S during the period of rewarming. To obtain an appropriate Ramsay score, we needed to increase the rate of administration of the drug in Group P, and to decrease this rate in Group M. After the drugs were discontinued, Group P required mechanical ventilation for less time. In conclusion, propofol is as effective as sufentanil or midazolam in controlling increased Vo(2) postoperatively. The initial dose of 2 mg.kg(-1).h(-1) had to be increased for most patients. In addition, propofol sedation is associated with a quicker recovery compared with midazolam and sufentanil. Implications: A prospective, randomized comparison of propofol, sufentanil and midazolam infusions revealed similar effects on hemodynamics, oxygen consumption, and rate of myocardial ischemia after aortic surgery, although propofol was associated with a quicker recovery compared with midazolam and sufentanil.