Severe aplastic anemia (SAA) has a poor prognosis in the absence of treatment. Current accepted therapeutic strategies include allogeneic stem-cell transplantation and immunosuppression, both resulting in long-term survival in the majority of patients. Although human leukocyte antigen (HLA)-matched sibling stem-cell transplantation is highly effective, the 25% probability of finding a suitable sibling donor within a family renders this approach available to only a minority of patients. Transplantation using HLA-matched, unrelated donors carries a high risk of treatment failure along with considerable toxicity. While combined immunosuppression with both antithymocyte globulin (ATG) and cyclosporine A (CSA) produces hematologic improvement in most patients, relapse is common. Late evolution of aplastic anemia to other serious hematologic disorders, including paroxysmal nocturnal hemoglobinuria (PNH), myelodysplasia, and acute leukemia, is also a significant problem following treatment with ATG/CSA. Recently, results of immunosuppression in SAA with another potent immunosuppressive agent, cyclophosphamide, were reported in a small number of patients. The overall response rate was similar to that seen with ATG/CSA, but relapse and late clonal disease were not observed during a long period of follow-up. A larger randomized trial comparing sustained hematologic response rates to either conventional immunosuppression with ATG/CSA or high-dose cyclophosphamide and CSA is now underway; secondary end points include response duration, event-free survival, and overall survival. Additionally, a number of protocols designed to test the efficacy of alternative immunosuppressive or immunomodulatory agents are being developed.