Background: Glofitamab is a T-cell engaging bispecific antibody (Ab) with a novel 2:1 configuration that confers bivalency for CD20 (B cells) and monovalency for CD3 (T cells). In a Phase I/II study (NCT03075696), escalating glofitamab doses were highly active and well tolerated in pts with R/R B-cell lymphomas, with obinutuzumab pretreatment (Gpt) and Cycle (C) 1 step-up dosing providing effective CRS mitigation. For the first time, we present pivotal Phase II expansion results in pts with R/R DLBCL and ≥2 prior therapies. Methods: All pts had DLBCL (DLBCL NOS, HGBCL, PMBCL, or trFL) and had received ≥2 prior regimens including ≥1 anti-(a) CD20 Ab and ≥1 anthracycline. IV Gpt (1000mg) was given 7 days before the first glofitamab dose. IV glofitamab was then given as step-up doses on Day (D) 1 (2.5mg) and D8 (10mg) of C1 and at the target dose (30mg) on D1 of C2–12 (21-day cycles). The primary endpoint was CR rate (best response during initial treatment) assessed by Independent Review Committee (IRC) using Lugano 2014 criteria. CRS was assessed using ASTCT criteria. Results: As of Sep 14, 2021, 107 pts had received ≥1 dose of study treatment (median age: 66 yrs [21–90]; Ann Arbor stage III–IV disease: 74%; IPI score ≥3: 54%; DLBCL NOS: 74%). Median prior therapies was 3 (2–7); 59% had ≥3 prior therapies and 35% had received prior CAR T-cells (CAR-Ts). Most pts were refractory to a prior aCD20 Ab-containing regimen (85%) and to their most recent regimen (85%). Many were refractory to their initial therapy (59%) and to prior CAR-Ts (32%). After a median follow-up of 9 months (0.1–16), ORR and CR rates by IRC were 50.0% and 35.2%, respectively. CR rates were consistent in pts with and without prior CAR-Ts (32% vs 37%). Median time to CR was 42 days (95% CI: 41–48). The majority of CRs (33/38; 87%) were ongoing at data cut. An estimated 84% of complete responders and 61% of responders remained in response at 9 months. At data cut, the projected 12-month OS rate was 48%, and 92% of complete responders were alive. These results are consistent with earlier Phase I data in 100 pts treated with target glofitamab doses ≥10mg (CR rate: 34%; estimated 20-month CR rate in complete responders: 72%). CRS occurred in 68% of pts, was primarily associated with the initial doses, and was mostly Gr 1 (51%) or Gr 2 (12%); Gr 3 (3%) and Gr 4 (2%) events were uncommon. All but 2 CRS events were resolved at data cut. Glofitamab-related neurologic AEs potentially consistent with ICANS occurred in 3 pts (all Gr 1–2). No glofitamab-related Gr 5 (fatal) AEs occurred. Glofitamab-related AEs leading to discontinuation were uncommon (3 pts, 3%). Conclusions: Fixed-duration glofitamab induces durable complete remissions and has favorable safety in pts with R/R DLBCL and ≥2 prior therapies, including those with prior exposure to CAR-Ts. Glofitamab is a promising new therapy for pts with heavily pretreated and/or highly refractory DLBCL. Clinical trial information: NCT03075696.