Sixty-nine consecutive patients (median age 54 years) were prospectively enrolled in a single-institution protocol for allogeneic transplantation with adjusted non-myeloablative fludarabine-melfalan-based conditioning including cyclosporin A and MMF, and one of three modes of serotherapy. Thirty-one donors (45%) were unrelated. The first cohort of 29 had ATG (Thymoglobulin 2 mg/kg  3 days), the subsequent 26 had Campath 30 mg  3 days subcutaneously, and the final cohort of 14 had 30 mg Campath once. The groups were similar as regards age, diagnosis and risk factors. Campath-patients had no acute toxicity, fewer days with fever and antibiotics, and required fewer transfusions than ATG-treated patients. 3-d-Campath patients showed lower lymphocyte counts from day þ 4, and CD4 þ , CD8 þ , CD19 þ and NK cells recovered slower than in ATG-treated patients. More Campath patients developed mixed chimerism that required DLI. 3-d-Campath induced more serious and opportunistic infections than ATG, which resulted in a greater non-relapse mortality and an impaired overall survival despite a low tumor-related mortality. The change of the Campath dosing schedule to one dose abrogated the deleterious effect of 3-d-Campath on immune recovery, severe infections and survival. Subcutaneous Campath is simple and provides strong immune suppression with no early toxicity, but dose limitation to 30 mg once is recommended.