Acute plasmapheresis (APP) is an additional tool for blood conservation during cardiac surgery. In a randomized study of 60 aortocoronary bypass patients undergoing APP, the influence of replacement of the withdrawn autologous plasma (10 mL/kg) by either colloids (low molecular weight hydroxyethyl starch solution [6% HES 200/0.5]) or crystalloids (Ringer's solution) was investigated. APP was performed by means of a centrifugation technique producing platelet-poor plasma. During and after cardiopulmonary bypass (CPB), either a cell saver (CS) or a hemofiltration (HF) device was also used for blood concentration. Almost three times as much crystalloid as HES solution was necessary for replacement of autologous plasma. Fluid balance during CPB was significantly more positive in the crystalloid patients, particularly when a CS was used. Blood loss was highest in the crystalloid patients in whom a CS was used in addition to APP, and these were the only patients who needed packed red cells. The platelet count, AT-III and fibrinogen plasma concentrations, colloid osmotic pressure, albumin, and total protein were significantly less compromised in the patients with colloid volume replacement. These parameters were closest to control values in patients receiving colloid replacement and HF. It is concluded that colloid is preferred for replacement of autologous plasma withdrawn by APP, and HF is superior to the CS when the combined technique for blood conservation is used.