BU-CY is the established non-TBI-based myeloablative conditioning regimen for allogeneic hematopoietic SCT. However, liver toxicity and hepatic veno-occlusive disease (VOD) are frequent life-threatening complications. Pharmacological considerations suggest that BU can trigger toxicity of subsequent CY. Recent animal data confirmed this hypothesis. Less liver toxicity and better outcomes were observed when mice were treated with the reversed order of CY and BU. We analyzed in this study liver toxicity and outcome in patients receiving BU-CY (16 patients) or CY-BU (59 patients). Liver function differed significantly with higher levels of liver function tests between day þ 10 and þ 30, and a higher cumulative incidence of VOD in the BU-CY cohort (2/16 (12.5%) vs 0/59 (0%), P ¼ 0.006). TRM was significantly higher in patients receiving BU-CY (cumulative incidence BU-CY 45%, CY-BU 17%, P ¼ 0.02), without yet translating into a significant survival difference (incidence for survival: BU-CY 38%, CY-BU 63%; hazard ratio 1.19 for BU-CY, 95% confidence interval 0.29-4.82, P ¼ 0.80). Rates of engraftment and relapse were not different. These data support the concepts derived from animal models in favor of CY-BU compared with traditional BU-CY and form the basis for prospective controlled comparisons.