PURPOSE: Open-label phase II study (RELATIVITY-060) to investigate the efficacy and safety of first-line nivolumab, a PD-1-blocking antibody, plus relatlimab, a lymphocyte-activation gene 3 (LAG-3)-blocking antibody, plus chemotherapy in patients with previously untreated advanced gastric cancer (GC) or gastroesophageal junction cancer (GEJC). METHODS: Patients with unresectable, locally advanced or metastatic GC/GEJC were randomly assigned 1:1 to nivolumab + relatlimab (fixed-dose combination) + chemotherapy or nivolumab + chemotherapy. The primary end point was objective response rate (ORR; per RECIST v1.1 by blinded independent central review [BICR]) in patients whose tumors had LAG-3 expression ≥1%. RESULTS: Of 274 patients, 138 were randomly assigned to nivolumab + relatlimab + chemotherapy and 136 to nivolumab + chemotherapy. Median follow-up was 11.9 months. In patients with LAG-3 expression ≥1%, BICR-assessed ORR (95% CI) was 48% (38 to 59) in the nivolumab + relatlimab + chemotherapy arm and 61% (51 to 71) in the nivolumab + chemotherapy arm; median progression-free survival (95% CI) by BICR was 7.0 months (5.8 to 8.4) versus 8.3 months (6.9 to 12.1; hazard ratio [HR], 1.41 [95% CI, 0.97 to 2.05]), and median overall survival (95% CI) was 13.5 months (11.9 to 19.1) versus 16.0 months (10.9 to not estimable; HR, 1.04 [95% CI, 0.70 to 1.54]), respectively. Grade 3 or 4 treatment-related adverse events (TRAEs) occurred in 69% and 61% of all treated patients, and 42% and 36% of patients discontinued because of any-grade TRAEs in the nivolumab + relatlimab + chemotherapy and nivolumab + chemotherapy arms, respectively. CONCLUSION: RELATIVITY-060 did not meet its primary end point of improved ORR in patients with LAG-3 expression ≥1% when relatlimab was added to nivolumab + chemotherapy compared with nivolumab + chemotherapy. Further studies are needed to address whether adding anti-LAG-3 to anti-PD-1 plus chemotherapy can benefit specific GC/GEJC patient subgroups.