Background: High-dose chemotherapy (HDT) followed by autologous stem cell transplant (SCT) remains a standard of care in patients (pts) with newly diagnosed (ND) multiple myeloma (MM). While lenalidomide (R) maintenance is acknowledged to improve outcomes, intensified consolidation (such as tandem-SCT) has yielded conflicting results. Allogeneic (allo) SCT holds the promise of curative potential at the cost of higher treatment-related mortality (TRM). In a previous phase 2 study, we showed a very low TRM rate (6.1%) and feasibility of 12 months (mos) of R maintenance (maint), with auto/allo SCT after R/adriamycin/dexamethasone (RAD). This prompted us to compare, on a randomized rather than a “biological assignment” basis, a second auto- versus (vs) an allo-SCT in pts with an unfavorable prognosis. Methods: The current protocol (DSMM XIV, NCT01685814) was set up according to a double 2x2-factorial design. Post-induction (PInd) CR rate was the efficacy endpoint for the comparison of RAD vs bortezomib (V)/RD (VRD; 3 cycles each). If pts had achieved >VGPR to HDT, a second randomization (2ndR) compared immediate R maint (arm A2) with a second auto-SCT (B2). In case of < VGPR, pts were randomized between a second auto- (C2) and allo-SCT (D2). Planned R maint. duration was 36 mos, except after allo (12 mos). Results: Between 05/2012-06/2016, 476 pts were randomized and 469 received at least one dose of study drug. Pts’ median age was 55 (range, 32–65) years. 11.3% of pts had FISH del17p; 11.6% had t(4;14); and 4.4% had t(14;16). PInd CR rate was 11.8% (90% CI, 7.9%-16.3%) with RAD and 13.0% (90% CI, 8.9-18.0) with VRD (P = .697). 382 pts underwent R2 with 279 pts. (73%) in >VGPR and 103 (27%) in < VGPR, respectively. Median duration of R maint (N = 298) was 21.2 mos for A2, 23.1 mos for B2, 27.4 mos for C2, and 11.0 mos. for D2. At a median follow-up of 40.2 (0.5-87.0) months, median PFS from first randomization with RAD was 41.7 (95% CI, 35.4-48.5) mos vs. 53.7 (95% CI, 46.2-63.1) mos with VRD (P = .0439). Median PFS from 2ndR was 38.7 (95% CI, 30.3-47.3) mos for the 181 RAD vs. 50.7 (95% CI, 44.4-64.9) mos for the 201 VRD pts (P = .0126). Median overall survival (OS) cannot be estimated. With 47 deceased RAD vs 36 VRD pts, HR was .671 (95% CI, .435-1.037; P = .0703). Conclusions: In this study, median PFS benefit was 12 mos in favor of VRD vs. RAD despite comparable PInd CR. We show for the first time a len-PI to be superior to a len-chemo triplet, confirmed with positive OS trends. 3-year PFS for all consolidation arms will be presented. Clinical trial information: NCT01685814.