Abstract High-dose chemotherapy followed by blood or marrow transplantation (BMT) is generally considered the best salvage option for patients with relapsed diffuse large–B-cell non-Hodgkin’s lymphoma (DLCL). The relative roles for allogeneic and autologous BMT remain controversial. We reviewed the clinical outcome of 183 patients with relapsed DLCL who underwent BMT at Johns Hopkins University in 1985-2001. A total of 45 patients received T-cell–depleted HLA-matched allogeneic bone marrow (allo-BMT), and 138 patients received autologous marrow or peripheral blood stem cells (auto-BMT). The allo-BMT recipients had a higher proportion of patients with chemoresistant disease ( P = .004) and had received more chemotherapy before BMT ( P = .02). The auto-BMT recipients were older ( P < .001) and were of more advanced-stage disease ( P = .01). The 3-year overall survival (OS) was 23.7% (median survival, 129 days) after allo-BMT and 33.1% (median survival, 263 days), after auto-BMT (log-rank, P = .17). The 3-year OS for patients with sensitive disease was 51.9% after allo-BMT and 46.2% after auto-BMT (log-rank, P = .38). For patients with resistant disease, the 3-year OS was 12.1% after allo-BMT and 19.1% after auto-BMT (log rank, P = .08). In multivariate analysis, significant predictors of death were disease sensitivity (hazard rate [HR], 0.3; 95% confidence interval [CI] 0.2-04; P < .001), age >40 years (HR, 2.42; 95% CI, 1.7-3.4; P < .001), and stage at diagnosis (HR, 1.2; 95% CI, 1.0-1.4; P = .04). The 3-year event-free survival (EFS) for patients with sensitive disease was 52.7% after allo-BMT and 42.0% after auto-BMT (log-rank, P = .29). For patients with resistant disease, the 3-year EFS was 6.2% after allo-BMT and 19.4% after auto-BMT (log-rank, P = .1). The 3-year probability of relapse for chemosensitive patients was 30% after allo-BMT and 46.1% after auto-BMT (log-rank, P = .25). The 3-year relapse rate in patients with resistant disease was 75.0% after allo-BMT and 69.9% after auto-BMT (log-rank, P = .58). In multivariate analysis, only disease sensitivity status (HR, 0.4; 95% CI, 0.2-2.1; P < .001) and age >40 years (HR, 1.7; 95% CI, 1.1-2.9; P = .03) appear to have a significant impact on relapse. Transplant-related mortality (TRM) was the cause of death for 51.1% of allo-BMT recipients and 23.9% of auto-BMT recipients ( P < .001). Mortality from lymphoma was 26.6% in allo-BMT recipients and 43.5% in auto-BMT recipients ( P = .02). Auto-BMT and allo-BMT produced similar survival for patients with relapsed DLCL. For patients with sensitive disease, allo-BMT seemed to provide longer survival with less relapse; however, this was achieved at the cost of greater TRM.