Allogeneic BMT provides the best treatment currently available for long-term disease-free survival in patients with recurrent ALL. Historically, partially matched related donors provided the opportunity for treatment to a greater number of patients than matched related donors at the expense of decreased overall survival. In this study we compare the results in recurrent ALL patients transplanted with either HLA identical sibling bone marrow or partially matched related bone marrow. Thirty-two patients with relapsed ALL received partially matched bone marrows from a relative with one to three HLA, A, B and Dr antigen mismatches. Bone marrow was partially T cell-depleted with murine T10B9.1A-31 moAb. Sixteen patients with relapsed ALL received HLA-matched sibling bone marrows. All partially matched patients received additional GVHD prophylaxis with methylprednisolone in addition to anti-CD5 immunotoxin and/or CYA. All matched patients in addition to methylprednisolone received MTX and/or CYA. We observed no difference in disease-free survival between patients transplanted with partially matched bone marrow (median follow-up 1252 days, range 778-2035 days) vs those transplanted with HLA-matched bone marrow (median follow-up 1472 days, range 1165-2800 days; P = 0.48). Median survival for all patients is 38% (95% CI 24-52%) at 6 years. Patients transplanted in remission had a significant increase in disease-free survival when compared to those in relapse (P = 0.007). Our data suggest that partially matched BMTs from related donors are a comparable alternative to fully matched transplants in patients with ALL.