Background: Adding Bev to erlotinib prolonged PFS in NEJ026 and CTONG1509 trials, but limited data are available adding Bev to a second-generation EGFR-tyrosine kinase inhibitor. AfaBev-CS is a Japanese no-profit, randomized, open-label, multicenter phase II trial of Afa plus Bev vs Afa alone as first-line treatment for EGFR-mutated advanced NSCLC. Methods: This study enrolled untreated pts of advanced non-squamous NSCLC harboring EGFR sensitizing mutation (Del19 or L858R) and without symptomatic brain metastases. 100 eligible pts were randomized in a 1:1 ratio to receive either Afa (30 mg, daily) plus Bev (15 mg/kg, every 3 weeks) (AfaBev arm) or Afa (40 mg, daily) monotherapy (Afa arm), and stratified according to stage, EGFR mutation status and institution. The primary endpoint was PFS and the secondary endpoints were OS, tumor response and time to treatment failure. The sample size was set in terms of feasibility. The power is greater than 50% under the assumptions of a median PFS of 12 months for the Afa arm and HR of 0.6 for the AfaBev arm, with an accrual of 2.5 years and a minimum planned follow-up period of 2 years with the type 1 error of 0.05 (two-sided). Results: Between August 2017 and September 2019, 100 pts were enrolled (each arm, 50 pts). At a median follow-up of 31.3 months for all randomized pts, total 69 events occurred. Median PFS was 16.3 months for AfaBev arm and 16.1 months for Afa arm, with a hazard ratio (HR) of 0.865 (95%CI, 0.539 – 1.388; loglank p = 0.5476). In subgroup analysis, pts < 70 years old (HR 0.347) and pts with brain metastasis (HR 0.353) showed better trend of PFS in AfaBev arm. On the other hand, pts ≥ 70 years old (HR 1.738) and pts without brain metastasis (HR 1.196) did not show better trend in AfaBev arm. In terms of OS, result was immature because number of events was still small. Objective response rate was 77.6% in AfaBev arm and 72.0% in Afa arm. Severe adverse events were observed in 11 pts for each arm. Grade 3 or more diarrhea, hypertension, rash acneiform, paronychia and stomatitis were frequently observed in AfaBev arm. Pneumonitis was not observed for AfaBev arm and observed in 3 (6.0%) for Afa arm, and grade 3 in one patient. Conclusions: This study failed to show the efficacy of AfaBev arm for improving PFS in untreated pts with EGFR mutated non-squamous NSCLC. In pts < 70 years old, Afa plus Bev might be promising. Clinical trial information: jRCTs061180006.