The performance of a leukocyte reduction bedside filter with different types of RBC concentrates was analyzed. Three types of RBCs were prepared; buffycoat-depleted RBCs suspended in saline-adenine-glucose-mannitol (SAGM)-additive solution (BC-RBCs; n = 20), RBCs suspended in SAGM-additive solution without huffy coat removal (SAGM-RBCs; n = 20), and RBCs drawn in CPDA-1 conservative solution and processed for component preparation by the platelet-rich plasma method (CPDA-RBCs; n = 20). The units were filtered within 8 h of collection. One filter was used for every 2 units. High numbers of residual WBCs were found even in the units filtered first. Filtration of CPDA-RBCs resulted in a higher residual WBC content than SAGM-RBCs or BC-RBCs(p = 0.0032 and p = 0.0002, respectively). The filter performance strikingly decreased when the WBC load per filter exceeded 4x 10^9 or the platelet load was less than 100 x 10^9. We conclude that filter performance varies with the WBC and platelet content of the RBC concentrates. Under the experimental conditions assayed in this study CPDA-RBCs are the least appropriate ones to be used for bedside leukocyte reduction.