Résumé : BACKGROUND: Perioperative fluid management - including the type, dose, and timing of administration -directly affects patient outcome after major surgery. The objective of fluid administration is to optimize intravascular fluid status to maintain adequate tissue perfusion. There is continuing controversy around the perioperative use of crystalloid versus colloid fluids. Unfortunately, the importance of fluid volume, which significantly influences the benefit-to-risk ratio of each chosen solution, has often been overlooked in this debate. MAIN TEXT: The volume of fluid administered during the perioperative period can influence the incidence and severity of postoperative complications. Regrettably, there is still huge variability in fluid administration practices, both intra-and inter-individual, among clinicians. Goal-directed fluid therapy (GDFT), aimed at optimizing flow-related variables, has been demonstrated to have some clinical benefit and has been recommended by multiple professional societies. However, this approach has failed to achieve widespread adoption. A closed-loop fluid administration system designed to assist anesthesia providers in consistently applying GDFT strategies has recently been developed and tested. Such an approach may change the crystalloid versus colloid debate. Because colloid solutions have a more profound effect on intravascular volume and longer plasma persistence, their use in this more "controlled" context could be associated with a lower fluid balance, and potentially improved patient outcome. Additionally, most studies that have assessed the impact of a GDFT strategy on the outcome of high-risk surgical patients have used hydroxyethyl starch (HES) solutions in their protocols. Some of these studies have demonstrated beneficial effects, while none of them has reported severe complications. CONCLUSIONS: The type and volume of fluid used for perioperative management need to be individualized according to the patient's hemodynamic status and clinical condition. The amount of fluid given should be guided by well-defined physiologic targets. Compliance with a predefined hemodynamic protocol may be optimized by using a computerized system. The type of fluid should also be individualized, as should any drug therapy, with careful consideration of timing and dose. It is our perspective that HES solutions remain a valid option for fluid therapy in the perioperative context because of their effects on blood volume and their reasonable benefit/risk profile.