To study the effects of M-CSF administration on long-term outcomes of unrelated BMT, we retrospectively analyzed data from patients transplanted through the Japan Marrow Donor Program. We obtained data from 54 patients who received M-CSF just after BMT and 500 patients who did not receive M-CSF or G-CSF acted as controls. There were no significant differences between the two cohorts with respect to OS, acute GVHD or relapse. Although the incidence of chronic GVHD was comparable between the two groups, extensive chronic GVHD was observed significantly less often in the M-CSF cohort than in the control group. Multivariate analysis identified M-CSF as a significant factor for attenuating extensive chronic GVHD (relative risk: 0.73; 95% confidence interval: 0.55-0.94; P=0.012). We also found the same results in matched-pair analysis. Our observation suggests the potential for clinical use of M-CSF to dampen severe chronic GVHD.