Transfusion-associated graft-versus-host disease (TA-GvHD) is caused by leukocytes, specifically T cells within a transfused blood product. Currently, the prevention of transfusion-associated graft-versus-host disease is performed by irradiation of blood products. With a sufficient reduction of leukocytes, the risk for TA-GvHD can be decreased. With consistent advances in current state-of-the-art blood filters, we herein propose that double filtration can sufficiently reduce leukocytes to reduce the risk for TA-GvHD. Thirty RBC concentrates were filtered with leukocyte filters, followed by storage at 1-6 oC for 72 hours, and then a second filtration was performed. Residual leukocytes in the double-filtered RBC units (n = 30) were assessed with flow cytometric methods, and an additional assay with isolated peripheral blood mononuclear cells (PBMCs) (n = 6) was done by both flow cytometric methods and an automated hematology analyzer. Quality of the RBCs after filtration was evaluated by hematological and biochemical tests. In vitro T cell expansion was performed using anti-CD3/CD28-coated Dynabeads or anti-CD3 (OKT3). In vivo experiment for GvHD was performed by using NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice. Double-filtered blood products showed residual leukocyte levels below detection limits, which calculated to be below 1200-2500 cells per blood unit. In vitro expansion rate of T cells showed that 6x103 and 1x103 cell-seeded specimens showed 60.8±10.6 fold and 10.2±9.7-fold expansion, respectively. Cell expansion was not sufficiently observed in wells planted with 1x102 or 10 cells. In vivo experiments showed that mice injected with 1x105 or more cells cause fatal GvHD. GvHD induced inflammation was observed in mice injected with 1x104 or more cells. No evidence of GvHD was found in mice injected with 103 cells. Our study suggests that additional removal of contaminating lymphocytes by a second leukodepletion step may further reduce the risk for TA-GvHD.