Allogeneic bone marrow transplantation (BMT) remains a curative treatment option for patients with CML. Unrelated donor (URD)-BMT is a realistic treatment option for a large group of patients (pts) in need. To ensure engraftment and to reduce the incidence of acute and chronic GVHD antithymocyte globulin (ATG) has been incorporated into the preparative regimen by many centers. However, little is known if the various preparations available yield comparable results. This analysis reports a single center experience on URD-BMT in chronic phase CML pts with special emphasis on the different ATG preparations used (immunization of rabbits with human thymus, ATG Merieux/Sangstat/Genzyme (ATG-M) or with the leukemic cell line Jurkat, ATG Fresenius (ATG-F)). Since 1995 61 pts (40 male; 21 female) with CML received an URD-BMT. Three pts (1 male, 2 female) from the ATG-F group were excluded because they also received ATG-M after early graft failures, which possibly were attributable to dose adjustments for obesity, to HHV6 infection, and prior use of Imatinib mesylate. Median patient age was 38 years (16–61). 46 pts were transplanted in first chronic phase and 12 in accelerated or second chronic phase and 76% received a fully matched (HLA-A, B, DRB1) transplant. 13 pts received a 1 antigen and 1 patient a 2 antigen mismatched transplant. All pts received in vivo T-cell depletion with ATG during conditioning. 46 pts received ATG-M and 12 pts ATG-F. As analyzed by July 2004 after a median follow-up of 3.3 years for surviving pts. the estimated probability for overall survival is 61%. For eventfree survival, despite the low number and short follow-up of pts who received ATG- F, there was a strong trend (p=0.0611; Log-Rank) in favor of ATG-F as compared to ATG-M, which reached significance in a matched pair analysis (p=0.03). In multivariate cox regression analysis for pretransplant risk assessment for overall survival a time interval of more than a year from diagnosis to transplant (RR 2.5, 1.0–6.3) and a CMV-IgG negative donor for a CMV-IgG positive patient (RR 2.9, 1.1–7.7) and for eventfree survival again the time interval from diagnosis to transplant (RR 2.3, 0.9–5.7) and the use of ATG-M (RR 4.7, 0.6–35.2) were independent negative predictors. The incidence of acute GVHD grade III - IV was higher in the ATG-M group (17% vs. 8%) as was the rate of extensive chronic GVHD (29% vs. 0%). This also translated into a higher risk of death without relapse (35% vs. 8%). However, these observations may be biased to some extent, since most patients who were receiving ATG-F received a conditioning regimen based on intravenous busulfan, whereas most patients who were receiving ATG-M received conditioning regimens mainly based on oral untargeted busulfan, although the type of conditioning was no independent risk factor in the Cox regression model used. Nevertheless, our data indicate that the results of URD-BMT in chronic phase CML patients can further be improved by reducing the rate of acute and chronic GVHD. A promising approach seems to be a conditioning regimen based on intravenous busulfan incorporating ATG-Fresenius.