Background: Primary mediastinal B-cell lymphoma (PMBCL) is a distinct entity of aggressive lymphoma, which typically presents in young patients (pts) with a bulky mediastinal mass. Therapy is based on R-CHOP or similar regimens, but the role of treatment intensification and consolidative radiotherapy (RT) is controversial, because data from randomized trials are rare. Methods: The UNFOLDER trial included 18-60 year-old pts (aaIPI = 0 with Bulk [≥7.5 cm] or aaIPI = 1) qualifying for radiotherapy to Bulk or extralymphatic involvement (E). Pts were randomized in a 2 x 2 factorial design to 6xR-CHOP-14 or 6x-R-CHOP-21 without RT or with RT (39.6 Gy) to Bulk and E. Primary endpoint was event-free survival (EFS), secondary endpoints were progression-free (PFS) and overall survival (OS). Response was evaluated by the Internat Standardized Response Criteria, Cheson 1999. Results: 131 PMBCLs were included with a median age of 34 years, 54% were female, 79% had elevated LDH > UNV and 24% had E. 82 pts (R-CHOP-21: 43; R-CHOP-14: 39) were assigned to RT and 49 (R-CHOP-21: 27, R-CHOP-14: 22) to no-RT. 96% (79/82) received RT per protocol and 5 pts in the no-RT arm received unplanned RT (4 after PR and 1 after CR/CRu). Response RT vs no-RT were CR/Cru 94% vs 84%, PR 2% vs 10%, PD 2% vs 4%. 3-year EFS was superior in pts assigned to RT (94% vs. 78%; p = 0.007), mostly due to events caused by initiation of RT (n = 5) in the no-RT arm. In an as treated analysis the difference between the RT and the no-RT arm was not significant (p = 0.136). Regarding PFS and OS no difference between the RT vs no-RT arm was detected (PFS: 95% (95% CI: 90-100) vs 90% (95% CI: 81-98), p = 0.253; OS: 98% (95% CI: 94-100) vs 96% (95% CI: 90-100), p = 0.636). Dose-densification of R-CHOP-21 by R-CHOP-14 did not improve EFS, PFS nor OS. Only 4 pts died. Conclusions: To our knowledge, this is the largest series of PMBCLs so far, which have been treated in a prospective, randomized trial in the rituximab era. The results reveal no differences between R-CHOP-14 vs R-CHOP-21. Pts assigned to RT had a superior EFS mostly due to a higher PR rate in the no-RT arm triggering RT, with no differences in PFS and OS. The results suggest a benefit of RT only for pts, who are responding to R-CHOP with PR. Testing RT in PET-positive residual tumors in a randomized trial can solve the question, while RT in PET-negative pts is studied in the ongoing randomized IELSG 37 trial. Our results indicate a very favorable 3-year OS of 96% in PMBCL pts treated with R-CHOP. Supported by Deutsche Krebshilfe, Amgen and Roche. Clinical trial information: NCT00278408.