The hemodynamic effects of two types of anesthesia on aortofemoral bypass surgery were studied in a randomised prospective trial. Epidural anesthesia supplemented with nitrous oxide (group I) and total intravenous anesthesia combining fentanyl and a continuous infusion of etomidate (group II) were compared. A high incidence of preoperative disease was found and all 18 patients were classified in ASA classes III-IV. It is concluded that epidural anesthesia provides excellent anesthetic and hemodynamic stability provided that an optimal filling pressure is maintained. Total intravenous anesthesia resulted in significant hypertensive reactions during surgery, which were not specifically related to crossclamping. Decreasing the high SVRI with vasodilatory treatment was necessary to treat hypertension in all those patients with preoperative hypertensive disease. No problems were seen in the intravenous group patients without preoperative hypertension. Cardiac work was higher in the intravenous group due to the high impedance of the cardiovascular system provoked by the absence of vasodilatory properties with this type of intravenous anesthesia. Monitoring of PWP and CI by Swan-Ganz catheter is shown to be very useful for optimalization of hemodynamics and fluid management especially during crossclamping, when normal Frank-Starling relationships might not be valid anymore. The effect of vasodilatory treatment, crossclamping and declamping could be carefully evaluated.