A consecutive prospective series of gastric cancer patients treated in our department from January 1982 to December 1990 is presented. There was complete follow-up (100%) for a minimum of 5 years or until death. Of the 286 patients, 222 (78%) were operated. Of these, 134 (60%) were resected; 91 with curative intent (68% of resected, 32% of all cases). The cancers were of the intestinal type (differentiated) in 63%, diffuse (undifferentiated) in 22%, and indeterminate in 15% (Laurèn classification). The tumor was located in the upper third, in the middle third and in the lower third in 33.6, 22 and 34.6%, respectively, and involved all of the stomach in 9.8% of the cases.The gastric and lymphatic resection was performed according to The General Rules for Gastric Cancer Study in Surgery and Pathology of the Japanese Research Society for Gastric Cancer (D<sub>2</sub> type of lymph node dissection). The operative mortality for curative resections was 6.5% (9.5% for total gastrectomy and 4.1% for subtotal gastrectomy). The operative mortality for total gastrectomy decrease from 14.3% in the first 4 years of the study to 7.1% in the last 5 years. Patients dying in the immediate postoperative period were not considered for the analysis of survival rates in operated cases. The 5-year survival rate was 12.2% for all cases (35/286). No patient without surgery or operated but not resected lived for 2 years. The 5-year survival rate in resected cases was 30% (35/118). 41.2% for curative resections (35/85) and 0% for palliative resections (0/33). The 5-year survival rate according to depth of penetration in the stomach wall for curative resections was 94.1% (16/17) in early gastric cancer (mucosal and submucosal limit), 44.4% (8/18) in tumors with muscular involvement, and 22% (11/50) with serosal extension. Early gastric cancer represented 19% (17/91) of the curative resections (6% of all cancers). The 5-year survival rate for all curative resections was 51% (24/47) for subtotal gastrectomies and 29% (11/38) for total gastrectomies. The distribution of early and advanced cancers in each group was 13/47 in subtotal gastrectomies and 4/38 in total gastrectomies. The 5-year survival rate for curative resections in advanced cancer (early gastric cancer excluded) was 35% (12/34) for subtotal gastrectomies and 21% (7/34) for total gastrectomies. In our community gastric cancer continues to be a highly lethal disease with an overall mortality from the disease or its therapy of 88% by 5 years, with the majority of cases presenting late, being only one third of all patients amenable to curative resections. However, long-term results have improved substantially since previous studies. The increased number of curative resections and its better survival rates, the increased number of early gastric cancers and an important decrease in operative mortality for total gastrectomies accounts for these results. Decreased survival rates in patients with cancers in the two upper thirds of the stomach, independent of the depth of the involvement, was seen.