Résumé : Background and objective We hypothesized that, in vascular surgery patients, the application of a goal-directed strategy based on a pulse contour-derived cardiac index would be associated with a better haemodynamic status than the application of routine perioperative care and that the amount of fluid and/or inotropes required in such a goal-directed therapy depended on the general anaesthetic technique used. Methods Patients undergoing peripheral arterial bypass grafting were randomly assigned to three groups. In group 1, haemodynamic management was performed according to routine clinical practice. In the two other groups (groups 2 and 3) a goal-directed therapy was applied aiming to maintain the pulse contour-derived cardiac index above 2.51 m -2min -1. Patients in groups 1 and 2 received sevoflurane-based anaesthesia and patients in group 3 propofol-based anaesthesia. Haemodynamic variables, amount of fluid and administration of inotropes were assessed at different time intervals. Results The amount of fluid administered was not significantly different between the groups. Two patients in group 1,13 patients in group 2 and 12 patients in group 3 were treated with dobutamine (P< 0.001). None of the patients anaesthetized with sevoflurane (groups 1 and 2) experienced postoperative cardiovascular complications, whereas four patients in the total intravenous group (group 3) experienced major postoperative cardiovascular complications (P = 0.005). Conclusion In the conditions of the present study, the application of a goal-directed therapy aiming to maintain the cardiac index above 2.5 l min -1 m -2 did not result in a higher tissue oxygen delivery than when applying the standard haemodynamic strategy nor did it depend on the anaesthetic technique used. © 2010 Copyright European Society of Anaesthesiology.