The results of leucocyte filtration during cardiac surgery are conflicting. This may be due to timing and duration of the filtration procedure, and to flow and pressure conditions in the filter. Therefore, we prospectively compared three major leucocyte filtration strategies in cardiac surgical patients. Forty patients were randomly divided into four groups. Group I: leucofiltration of arterial blood throughout cardiopulmonary bypass (CPB) (associated with high-flow and pressure gradients), Group II: leucofiltration of a part of the venous return blood in the rewarming phase during CPB (associated with intermediate flow, but high pressure), Group III: leucofiltration of residual heart-lung machine blood during transfusion into the patient after CPB (associated with low flow and low pressure), Group IV: control group without leucofiltration. We measured circulating leucocyte counts, plasma elastase levels and arterial blood oxygenation. Filters were postoperatively examined using scanning electronmicroscopy (SEM). Leucocyte counts increased over time and oxygenation decreased in all groups, without significant differences between the groups. SEM demonstrated extensive protein deposits and damaged leucocytes in the deeper layers of the filters from Group I. This was not observed in the filters from Group III. The postoperative plasma elastase levels increased in Groups II and IV and decreased in Groups I and III. In conclusion, we could not demonstrate a clinical difference among the three leucocyte depletion strategies. However, our laboratory results suggest that leucocyte filtration at low flow and pressure conditions is associated with less leucocyte damage and less release of elastase.