BACKGROUND: Adding pertuzumab to trastuzumab and chemotherapy improves survival in HER2-positive early breast cancer and metastatic breast cancer. We assessed the efficacy and safety of pertuzumab versus placebo in combination with trastuzumab and chemotherapy in first-line HER2-positive metastatic gastric or gastro-oesophageal junction cancer. METHODS: JACOB was a double-blind, placebo-controlled, randomised, multicentre, phase 3 trial in patients aged 18 years or older with HER2-positive metastatic gastric or gastro-oesophageal junction cancer. Eligible patients had measurable or evaluable non-measurable disease at baseline, Eastern Cooperative Oncology Group performance status of 0 or 1, and baseline left ventricular ejection fraction of 55% or more. Patients at 197 oncology clinics (in 30 countries) were randomly assigned (1:1) to receive either pertuzumab (840 mg intravenously) or placebo every 3 weeks, with trastuzumab (8 mg/kg loading dose, then 6 mg/kg every 3 weeks intravenously), plus chemotherapy (cisplatin 80 mg/m every 3 weeks intravenously, oral capecitabine 1000 mg/m twice a day [2000 mg/m every 24 h] for 28 doses every 3 weeks, or 5-fluorouracil 800 mg/m every 24 h intravenously [120 h continuous infusion] every 3 weeks). Randomisation was by a central permuted block randomisation scheme (block size of 4) with an interactive voice or web response system, stratified by geographical region, previous gastrectomy, and HER2 positivity. The primary endpoint was overall survival in the intention-to-treat population. This trial is registered with Clinicaltrials.gov, number NCT01774786 (ongoing, but closed to enrolment). FINDINGS: Between June 10, 2013, and Jan 12, 2016, of 3287 patients assessed, 780 eligible patients were randomly assigned to receive either pertuzumab plus trastuzumab and chemotherapy (pertuzumab group, n=388) or placebo plus trastuzumab and chemotherapy (control group, n=392). Median duration of follow-up was 24·4 months (95% CI 22·3-26·1) in the pertuzumab group and 25·0 months (22·3-28·9) in the control group. After 242 deaths in the pertuzumab group and 262 deaths in the control group (the study was not stopped at this point), overall survival was not significantly different between treatment groups (median overall survival 17·5 months [95% CI 16·2-19·3] in the pertuzumab group and 14·2 months [12·9-15·5] in the control group; hazard ratio 0·84 [95% CI 0·71-1·00]; p=0·057). Serious adverse events occurred in 175 (45%) of 385 patients in the pertuzumab group and 152 (39%) of 388 patients in the control group. Diarrhoea was the most common serious adverse event in both groups (17 [4%] patients in the pertuzumab group vs 20 [5%] patients in the control group). The most common grade 3-5 adverse events were neutropenia (116 [30%] patients in the pertuzumab group vs 108 [28%] patients in the control group), anaemia (56 [15%] vs 65 [17%]), and diarrhoea (51 [13%] vs 25 [6%]). Treatment-related deaths occurred in seven (2%) patients in the control group; no treatment-related deaths occurred in the pertuzumab group. INTERPRETATION: Adding pertuzumab to trastuzumab and chemotherapy did not significantly improve overall survival in patients with HER2-positive metastatic gastric or gastro-oesophageal junction cancer compared with placebo. Further studies are needed to identify improved first-line treatment options in these types of cancer and to identify patients with HER2-driven tumours who might benefit from dual HER2-targeted therapy. FUNDING: F. Hoffmann-La Roche Ltd.