Background: We investigated the superiority of docetaxel plus cisplatin and S-1 (DCS) compared with cisplatin plus S-1 (CS) in chemotherapy-naive patients with advanced gastric cancer. Methods: This randomised, open-label, phase 3 trial was conducted at 56 hospitals in Japan. Patients were randomised (1:1) to receive DCS (docetaxel 40 mg/m2 and cisplatin 60 mg/m2 on day 1, and S-1 40-60 mg twice a day for 2 weeks, every 4 weeks) or CS (cisplatin 60 mg/m2 on day 8, and S-1 40-60 mg twice a day for 3 weeks, every 5 weeks). The primary endpoint was overall survival in the intention-to-treat population. The study is registered with UMIN-CTR, number UMI07652. Findings: Between April 3, 2012, and March 18, 2016, 741 patients were randomised to DCS (n=370) or CS (n=371). Median overall survival was 14·2 months (95% CI 12·9-15·9) in the DCS group and 15·3 months (14·2-16·2) in the CS group (hazard ratio [HR] 0·99, 95% CI 0·85-1·16; one-sided stratified log-rank p=0·47). Median progression-free survival was 7·4 months for DCS and 6·5 months for CS (HR 0·99; 95% CI, 0·86-1·15; p=0·92), and the overall response rate was 59% for DCS and 56% for CS (p=0·50). In the safety analysis set, the most common grade 3 or worse adverse events were neutropenia (209 [58%] of 358 patients in the DCS group vs 117 [32%] of 367 patients in the CS group), leukopenia (120 [34%] vs 60 [16%]), and anorexia (94 [26%] vs 81 [22%]). Interpretation: DCS did not improve overall survival in chemotherapy-naive Japanese patients with advanced gastric cancer. Therefore, CS remains the standard first-line chemotherapy. Trial Registration Number: The study is registered with UMIN-CTR, number UMI07652. Funding: Ministry of Health, Labour and Welfare, and Japan Agency for Medical Research and Development. Declaration of Interest: YY reports personal fees from Taiho, Chugai, Eli Lilly (EL), Yakult, Nipponkayaku, Ono, Eisai, Bristol-Myers Squibb (BMS), Janssen, 1 AMED and research grant from Daiichi-Sankyo (DS). NB has received grants and honoraria from Taiho and BMS, and honoraria from Ono, Chugai, EL. JM, HF, and SK reports grants and non-financial support from MHLW and AMED, and SK personal fees from Johnson & Johnson (J&J). SI reports personal fees and others from EL and Chugai, personal fees from Taiho and Ono, others from BMS, Eisai, DS, Novartis, Merck-Serono (MS), Bayer, Otsuka. SK reports grants and personal fees from EL, Taiho, Ono, BMS, grants from Boehringer-Ingelheim (BI), personal fees from Chugai, MS, Bayer, Eisai, and Yakult. KS reports grants and personal fees from EL, Ono, Novartis, Yakult, personal fees from Astellas, BMS, Takeda, Pfizer, Novartis, and Abbvie, grants from Dainippon-Sumitomo, DS, Taiho, Chugai, and MSD. TT reports grants from AMED, Ono, BMS, Takeda, personal fees from MS and Takeda. HH reports grants from AstraZeneca, DS, Dainippon-Sumitomo, EL, MS, MSD, Taiho, Chugai, Eisai, LSK BioPharma, Incyte, Pfizer, BI, Beigene, Ono, and BMS. TY reports personal fees from Ono, BMS, EL, Taiho, Chugai, Yakult, DS, NipponKayaku, Olympus, Covidien, J&J, and Terumo. MS reports personal fees from Taiho, Chugai, EL, Ono, DS, Yakult, Merck Sharp and Dohme. MT reports personal fees from Taiho, Chugai, Ono, BMS, Yakult, Takeda, EL, Pfizer, and DS. NN, MA, KC, HH, MN, and HY have nothing to disclose. Ethical Approval: The trial protocol was approved by the JCOG Protocol Review Committee and the institutional review boards of each participating centre.