Background: It is uncertain if the addition of oxaliplatin (OX) to fluoropyrimidine plus bevacizumab (BEV) is suitable as initial therapy in elderly patients (pts) with metastatic colorectal cancer (MCRC). Therefore, we conducted a randomized controlled trial to confirm the superiority of the addition of OX in terms of progression-free survival (PFS). This JCOG trial was originally planned as a paralell study with NCCTG, but the NCCTG trial was terminated early. Methods: Key eligibility criteria included unresectable metastatic colorectal cancer, and histologically confirmed adenocarcinoma, aged 70-74 with PS 2 or 75 or older with ECOG PS 0-2. Eligible pts were randomized (1:1) to either no addition of OX or addition. Whether using 5-FU+levoleucovorin calcium (5-FU/l-LV) or capecitabine (CAPE) was declared before study entry; options included 5-FU/l-LV+BEV (C), CAPE+BEV (D), mFOLFOX7+BEV (E), or CapeOX+BEV (F). 5-FU/l-LV regimen omitted bolus 5-FU from the original sLV5FU regimen. The dose of CAPE was adjusted by estimated creatinine clearance. The primary endpoint was PFS. The planned sample size was 250 pts in total to detect a hazard ratio (HR) of 0.75, with a one-sided alpha of 5% and 70% power. The decision rule is that the primary endpoint is met, and the point estimate of HR of overall survival (OS) is less than 0.8. Results: Between Sep 2012 and Mar 2019, 251 pts were randomized. 125 pts were allocated to no addition of OX and 126 pts to addition. Median age was 79, aged 70-74/75-79/80-84/85+:5%/45%/37%/13%, PS 0/1/2:53%/39%/7%. Of 251 pts, 241 pts had PFS events and 223 pts had OS events. Median PFS (mPFS) was 9.4 months (M) (95%CI 8.3–10.3) in no addition of OX and 10.0M (9.0–11.2) in addition (HR 0.837, 90.5%CI [0.673–1.042], one-sided p = 0.086). Median OS was 21.3M (18.7-24.3) in no addition and 19.7M (15.5–25.5) in addition of OX (HR 1.054 [0.810–1.372]). Response rate was 29.5% (21.2-38.8) in no addition of OX and 47.7% (38.1-57.5) in addition. Proportion of pts whose EQ-5D scores improved from baseline to post-treatment in overall score did not differ (odds ratio 0.94 (0.51-1.75)). The deaths of 1 pt in no addition of OX and in 3 pts in addition were deemed treatment-related. Conclusions: The addition of OX has no survival benefit over no addition. OX was not recommended for elderly MCRC pts as initial therapy. Clinical trial information: UMIN000008866.