OBJECTIVE To assess the immunosuppressive capacity of hyperfractionated total lymphoid irradiation and cyclophosphamide for transplantation of unmodified allogeneic marrow in sensitized aplastic anemia patients. METHODS From February 1983 to September 1990, 23 multiply transfused aplastic anemia patients underwent unmodified bone marrow transplantation from HLA genotypically identical sibling donors following preparation with 6 Gy hyperfractionated total lymphoid irradiation and 160 mg/kg cyclophosphamide. Graft-versus-host disease prophylaxis included steroids in one patient, methotrexate in four, cyclosporine in seven, and methotrexate/cyclosporine in 12. There were 17 males and 6 females with a median age of 13 (range: 2.5-32). RESULTS One patient died early before engraftment of bacterial sepsis. Twenty-two patients were evaluable for engraftment. Three experienced graft failure including one primary, and two late graft failures associated with cyclosporine withdrawal. Acute graft-versus-host disease occurred in 7/22 (> or = grade II in 6), and chronic graft-versus-host disease in 3/17 patients. Except for a patient who received total body irradiation for a second transplant, no patient in this series developed interstitial pneumonia. Fifteen patients are alive with follow-up of 38-125 months (median 68). The overall actuarial survival at 5 years is 69%, at 8 years it is 60%, with one late death. The survival of adult patients was similar to that of younger patients (> or = 16 years old: 63%, < 16 years old: 55%). The development of acute graft-versus-host disease adversely influenced survival (88% with Grade 0-I, 17% with grade II-IV; p = 0.002). No hypothyroidism or secondary malignancies have been documented in this series. CONCLUSIONS Pretransplant immunosuppression with 6 Gy of hyperfractionated total lymphoid irradiation and 160 mg/kg CY reduces but does not eliminate the incidence of graft failure in sensitized aplastic anemia patients. The dose and the mode of administration of total lymphoid irradiation in this trial may be associated with a lower incidence of late side effects. Survival is comparable to that obtained using preparative regimens without radiation.