Acute and chronic GVHD contribute to much of the morbidity and mortality associated with unrelated bone marrow transplantation (UBMT). Effective prevention of these complications is crucial for the success of UBMT. Cyclosporine(CSA) was introduced for the prophylaxis of GVHD and has been shown to be effective; similar findings have been shown for tacrolimus. However, there are no comparative studies between tacrolimus/MTX and CSA/MTX in patients with SAA who undergo UBMT. We compared tacrolimus/MTX with CSA/MTX in patients with SAA who received UBMT using a matched-pair analysis. Patients with acquired SAA who received UBMT between July 1997 and December 2002 through the Japan Marrow Donor Program were eligible for analysis. Recipients of UBMT were divided into two cohorts for comparison: those who received tacrolimus/MTX and those who received CSA/MTX. Because our previous studies identified that recipient age and conditioning regimen are the most important variables associated with treatment failure, we selected pairs of patients treated with either tacrolimus/MTX or CSA/MTX by matching these variables. We were able to identify 47 pairs that could be matched exactly for the recipient age and conditioning regimens. There was an imbalance of HLA-A, -B, and -DRB1 antigen mismatches, with 18 mismatched pairs in the tacrolimus group and 7 in the CSA group (p=0.005). Other variables were comparable between the two groups. Engraftment took place in 46 patients (98%) in the tacrolimus group and 43 patients (92%) in the CSA group. The median time to neutrophil recovery was the same in the two groups (tacrolimus group: 18.0 days, CSA group: 17.5 days). The cumulative incidence of grade 2 to 4 acute GVHD was 28.9% in the tacrolimus group and 32.6% in the CSA group (p=0.56). The estimate of chronic GVHD occurrence was 13.3% for the tacrolimus group and 36.0% for the CSA group (p=0.10). 8 patients died of transplantation related toxicities in the tacrolimus group and 22 patients in the CSA group (p=0.002). There was a higher frequency of deaths from thrombomicroangiopathy in the CSA group than in the tacrolimus group (6 vs 0: p=0.03). The 4-year survival rate was 82% for patients in the tacrolimus group and 52% for patients in the CSA group (p=0.001). Although a disadvantage existed in terms of HLA disparity in the tacrolimus group, there was no difference in the incidence of acute and chronic GVHD between the two groups. The present study shows the superiority of tacrolimus/MTX over CSA/MTX in overall survival because of lower incidence of regimen-related toxic deaths.