OBJECTIVE We have reviewed the results of bone marrow transplantation in 30 patients with thalassemia major who were treated in the United States. METHODS Ten patients who underwent transplantation in Seattle and 20 patients from five other U.S. centers were identified through a survey of the International Bone Marrow Transplant Registry. These transplants were performed between November 1981 and April 1992 in patients with diverse ethnic backgrounds and ranged in age from 6 months to 14 years (median 4.0 years). Twenty-seven of the 30 patients received marrow from a human leukocyte antigen (HLA)-identical sibling or other family member, one patient received HLA-matched marrow from an unrelated donor, and two patients were given haploidentical but HLA-mismatched marrow from a related donor. Cytoreductive (preparative) therapy varied among institutions and pretransplant risk categories. In general, patients were given busulfan (12-24 mg/kg) or dimethylmyleran (5 mg/kg) in combination with cyclophosphamide (120-240 mg/kg). A subset of patients were given total body irradiation (TBI) at a dose of 720 cGy followed by cyclophosphamide (120 mg/kg). RESULTS Sixteen of 27 patients (59%) who received marrow from an HLA-identical family member are event-free survivors, with a duration of follow-up ranging from 2 months to > 10 years after transplantation. Six of these 27 patients (22%) had recurrence of thalassemia and five (19%) died. The estimated actuarial rate of thalassemia recurrence was 24% and the rate of event-free survival was 57%. Only one of the three patients who received marrow from HLA-nonidentical or unrelated donors survives event-free. Liver biopsies were not routinely performed before transplant. Thus, classification of patients into Lucarelli risk groups was not possible. A modified risk classification was devised by using liver size and iron status assessed by the regularity of chelation and the serum ferritin level. With use of this classification, there was no significant difference in event-free survival between transplant risk groups. CONCLUSIONS The findings observed in this small series of patients confirms that thalassemia can be cured with bone marrow transplantation. Although most patients are event-free survivors, a significant number experienced recurrence of their disease. A cooperative multicenter trial of U.S. transplant centers may be necessary to evaluate the use of marrow transplantation for thalassemia and to determine optimal treatment.