First-line therapy for younger patients with peripheral T-cell non-Hodgkin lymphoma (T-NHL) consists of 6 courses of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) with or without etoposide (CHOEP), consolidated by high-dose therapy and autologous stem cell transplantation (auto-SCT). We hypothesized that allogeneic stem cell transplantation (allo-SCT) could improve outcomes. 104 patients with peripheral T-cell non-Hodgkin lymphoma, except ALK+ anaplastic large cell lymphoma, 18 to 60 years, all stages, and all age adjusted International Prognostic Index scores, except 0 and stage I, were randomized to 4 cycles of CHOEP and 1 cycle of dexamethasone, cytosine-arabinoside, and platinum (DHAP) followed by high-dose therapy and auto-SCT or myeloablative conditioning and allo-SCT. The primary end point was event-free survival (EFS) at 3 years. After a median follow-up of 42 months, the 3-year EFS after allo-SCT was 43%, as compared with 38% after auto-SCT. Overall survival at 3 years was 57% vs 70% after allo- or auto-SCT, without significant differences between treatment arms. None of the 21 responding patients proceeding to allo-SCT relapsed, as opposed to 13 of 36 patients (36%) proceeding to auto-SCT. Eight of 26 patients (31%) and none of 41 patients died of transplant-related toxicity after allo- and auto-SCT, respectively. The strong graft-versus-lymphoma effect after allo-SCT was counterbalanced by transplant-related mortality. This trial is registered at www.clinicaltrials.gov as #NCT00984412.