Although peritoneal lavage cytology is widely performed during surgery for gastric cancer and the results have been reported to be one of the accurate prognostic factors, the cancer stage is determined independent of the results of lavage cytology according to the First English Edition of Japanese Classification of Gastric Carcinoma. In this study we demonstrated the validity of lavage cytology for accurately staging gastric cancer. Between 1988 and 1996, peritoneal lavage cytology was performed in 347 patients with resectable gastric cancer. Among them, cytology was positive in 29 cases (8.4%). The survival rate of the cytology-positive patients in each stage was worse than that of all patients in the same stage. The prognosis of patients with positive cytology findings and serosa-exposed gastric cancer was significantly worse than that of negative cytology findings and serosa-exposed gastric cancer, and similar to that of negative cytology findings and serosa-infiltrating gastric cancer. Our data indicated that positive cytology findings thus indicated a poor prognosis, and the prognostic difference between positive and negative cytology findings was approximately a one-stage difference in the Japanese stage grouping. Based on our findings, the results of peritoneal lavage cytology should thus be included in the factors for staging gastric cancer.