Platelet transfusion effectiveness may be limited in multiply transfused patients by the development of the refractory state. White cell (WBC)-reduction filters with variable efficiency (1–3 log10 reduction) are available and have been shown to be effective in reducing the incidence of platelet alloimmunization. However, the threshold number of WBCs below which alloimmunization would no longer occur is yet to be determined. A previously established animal model was used to examine the relative efficiency of second- and third- generation filters in reducing the frequency of refractoriness to allogeneic platelets. In this model, California Black rabbits are used as blood donors and New Zealand White rabbits as transfusion recipients. Eight weekly transfusions of either second-generation or third-generation WBC-reduced blood resulted in no difference between the two groups in mean platelet survival and rate of refractoriness to allogeneic platelets. To evaluate the possible incremental benefit of removing supernatant plasma to prevent platelet refractoriness, experiments were performed in which groups of animals were given transfusion(s) with red cell suspensions that had been WBC-reduced or both plasma-depleted and WBC-reduced. A significantly lower rate of allogeneic platelet refractoriness was seen in the rabbits that received WBC-reduced and plasma-depleted red cells than in those that received red cells that had been WBC-reduced only. These data provide evidence that the combined use of plasma depletion and WBC reduction can decrease still further the frequency of refractoriness produced by allogeneic blood transfusions.