Recently, a less toxic regimen comprising reduced cyclophosphamide (Cy), fludarabine, and anti-thymocyte globulin (ATG) (Cy-Flu-ATG) was used to condition high-risk patients scheduled for allogeneic hematopoietic cell transplantation (alloHSCT) instead of standard Cy-ATG in patients with severe aplastic anemia (AA). We performed a randomized phase III study to compare the regimen-related toxicities (RRTs) of two different conditioning regimens: Cy-ATG vs. Cy-Flu-ATG. Patients in the Cy-ATG arm received Cy at 200 mg/kg. Those in the Cy-Flu-ATG arm received fludarabine (Flu) at 150 mg/m(2) and Cy at 100 mg/kg. A total of 83 patients (40 in the Cy-ATG and 43 in the Cy-Flu-ATG) were enrolled. Seventy-nine patients had AA and four had MDS. All predefined RRTs were significantly lower in patients of the Cy-Flu-ATG arm (23.3 vs. 55.0 %; p = 0.003). Infection with identified causative organism and sinusoidal obstruction syndrome, hematuria, febrile episodes, and death from any cause tended to be more frequent in Cy-ATG arm but did not differ significantly between arms. There was no difference in neutrophil engraftment failure (2.5 vs. 2.33 %; p = 0.959), acute graft-versus-host disease (GvHD) (15.0 vs. 23.3 %; p = 0.388), and chronic GvHD (16.7 vs. 16.2 %; p = 0.961) between Cy-ATG and Cy-Flu-ATG arms. The 4-year survival rate did not differ between the Cy-ATG and Cy-Flu-ATG arms. Preconditioning with Cy-Flu-ATG was superior to that afforded by Cy-ATG in terms of reducing RRT levels without increasing engraftment failure. (ClinicalTrials.gov number: NCT01145976).