Objective: Owing to the risks of serious and sustained toxicity, anticancer drugs such as cisplatin and irinotecan cannot be readily administered to patients with gastric cancer and severe peritoneal metastasis. Therefore, a standard chemotherapy regimen has yet to be established for these types of patients. This randomized study investigated the utility of sequential methotrexate and 5-fluorouracil therapy vs. 5-fluorouracil continuous infusion for gastric cancer with peritoneal metastasis. Methods: Eligible patients had radiologically confirmed peritoneal metastasis with intestinal stenosis, peritoneal tumor or ascites. Treatment with 5-fluorouracil continuous infusion (800 mg/m 2 /day, ci, d1-5, q4w) or methotrexate and 5-fluorouracil therapy (methotrexate, 100 mg/m 2 , bolus infusion, followed 3 h later by 5-fluorouracil, 600 mg/m 2 , bolus infusion, with leucovorin rescue, q1w) was continued until disease progression or unacceptable toxicity. The projected sample size was 236, providing 80% power to detect a 40% increase in median overall survival in methotrexate and 5-fluorouracil therapy with a one-sided a of 0.05. Results: All 237 randomized patients were included in the primary analysis. The methotrexate and 5-fluorouracil therapy arm was not superior to the 5-fluorouracil continuous infusion arm (median survival time, 9.4 months in the 5-fluorouracil continuous infusion arm, 10.6 months in the methotrexate and 5-fluorouracil therapy arm; hazard ratio, 0.94; 95% confidence interval, 0.72 -1.22; one-sided P ¼ 0.31). Frequencies of Grade 3 or higher neutropenia, Grade 3 or higher anorexia and treatment-related deaths were 0.9, 27.4 and 1.7%, respectively, in the 5-fluorouracil continuous infusion arm, and 31.9, 33.6 and 0.9%, respectively, in the methotrexate and 5-fluorouracil therapy arm. Conclusions: Methotrexate and 5-fluorouracil therapy is not suitable for use as standard therapy for advanced gastric cancer with peritoneal metastasis.