Both immunosuppressive therapy and bone marrow transplantation are accepted treatments for patients with aplastic anemia. Choosing one of these therapies for a given patient depends not only on donor availability but also on such factors as patient age. To compare survival rates and long-term complications after bone marrow transplantation or immuno-suppressive therapy in patients with acquired aplastic anemia and to identify prognostic factors associated with improved survival. Center-based, retrospective analysis. Referral center for patients with aplastic anemia. 395 patients with acquired aplastic anemia. Bone marrow transplant from an HLA-identical, related donor or immunosuppressive therapy. Kaplan-Meier survival curves, results of log-rank tests, and cumulative incidence curves. Of 168 bone marrow transplant recipients, 89% had sustained engraftment. Forty-six patients developed grade II to IV acute graft-versus-host disease, and 68 developed chronic graft-versus-host disease that required therapy. Of 227 patients who received immunosuppressive therapy, 44% achieved a complete, partial, or minimal response. Fifty-four percent died or had no response to therapy. Actuarial survival at 15 years was 69% for bone marrow transplant recipients and 38% for patients receiving immunosuppressive therapy (P < 0.001). Improved survival was associated with having bone marrow transplantation as primary therapy, being younger, having no transfusion before transplantation, and having a higher absolute neutrophil count. Disease duration, year of therapy, sex, refractoriness to platelet transfusions, and previous treatment with androgens or corticosteroids did not significantly affect survival. Data from this center suggest that bone marrow transplantation may be preferred for younger patients with acquired aplastic anemia who have matched, related donors. Long-term survival is excellent for patients who respond to either form of therapy.