Objectives Three leucoreduction filters were evaluated – when used alone or combined with centrifuge leucoreduction (C‐LR) – to prevent alloimmune platelet refractoriness in a dog platelet transfusion model. Materials and Methods Donor platelet‐rich plasma (PRP) or buffy coat (BC) platelets were either filter leucoreduced (F‐LR) or F‐LR/C‐LR, 51Cr radiolabelled and transfused. Weekly transfusions were given for up to 8 weeks or until platelet refractoriness. Recipients who accepted treated transfusions were then given non‐leucoreduced (non‐LR) platelets to determine whether donor‐specific tolerance had been induced. Results Acceptance of F‐LR PRP transfusions ranged from 29% to 66%. F‐LR/C‐LR transfusions prepared from PRP were accepted by 92%, from BC by 63% and from pooled PRP by 75% of recipients (p=NS); overall acceptance rate of F‐LR/C‐LR transfusions was 83%. Tolerance to subsequent non‐LR transfusions occurred in 45% of the F‐LR‐/C‐LR‐accepting recipients unrelated to DR‐B compatibility between donors and recipients (P = 0·18). Conclusion In a dog platelet transfusion model, acceptance of donor platelets required combining F‐LR with C‐LR as apparently each process removes different immunizing WBCs.