We report the preliminary results of a prospective randomized study on the impact of two different dosages of Cyclosporine A (Cs-A) on probability of development of acute and chronic GVHD, transplant-related mortality (TRM), relapse rate (RR) and event-free survival (EFS). Fifty-nine pediatric patients given BMT from an HLA-identical sibling were centrally randomized to receive either Cs-A at a dosage of 1 mg/kg/die (CsA1) or at a dosage of 3 mg/kg/die (CsA3) intravenously for the first 21 days after BMT. Patients given Cs-A at a dosage of 1 mg/kg/die had a higher probability of developing acute GVHD, but a lower relapse rate, which translated into a better probability of EFS. These preliminary results to be confirmed with a longer follow-up suggest that the use of low doses of CsA is feasible even though associated with a higher incidence of GVHD, but without any increment in TRM. The reduction of immunosuppressive treatment after BMT favoured the development of a graft-versus-leukemia effect, which seems to play a relevant role in preventing leukemia recurrence and in improving the cure rate.