OBJECTIVE To evaluate the effects on hemostasis of three different plasma substitutes with special reference to a newly developed hydroxyethyl starch used as priming solution in an extracorporeal circuit as well as peri- and postoperative infusion fluid, we studied 36 patients randomly assigned to one of three groups, undergoing coronary artery bypass grafting. METHODS The compositions of the priming solutions were: 2.5% hydroxyethyl starch; 3% gelatin; and 4% human albumin. Platelet function tests and clotting assays were performed on blood samples collected during and after cardiopulmonary bypass. RESULTS We found that plasma von Willebrand Factor remained higher in the human albumin group. Hydroxyethyl starch preserved platelet agglutination as well as human albumin, whereas platelet aggregation induced by adenosine 5'-di phosphate (ADP) proved to be similarly affected during cardiopulmonary bypass in the three study groups. Prolongation of the in vitro bleeding constant during the bypass period and subsequent partial recovery showed an affected platelet function in all groups during cardiopulmonary bypass. The clotting times, activated partial thromboplastin time and prothrombin time were similar in the three groups. Blood loss, peri- and postoperatively, showed also no differences. Hydroxyethyl starch appeared most cost-effective as priming solution in an extracorporeal circuit. CONCLUSIONS We conclude that, with human albumin the golden standard, 2.5% hydroxyethyl starch is a suitable colloid plasma substitute to be used as priming solution in an extracorporeal circuit as well as peri- and postoperative infusion fluid, reasonably well maintaining hemostasis.