There has been no apparent improvement in overall salvage of patients with cancer of the stomach treated by elective extended total gastrectomy from 1950-1964 as compared with those treated by partial gastrectomy during the preceding 20-year period, 1931-1950 at Memorial Hospital in New York City. Criteria of resectability have been extended in recent years, and, therefore, the two series of patients cannot be considered strictly comparable. In the 94 patients subjected to elective total gastrectomy for cancer of the stomach, more than half (55 patients) had cancer in the proximal 1/3 of the stomach. The results obtained in this group by total gastrectomy are inferior to those obtained in the earlier series by partial gastrectomy. Patients with carcinoma of the mid 1/3 of the stomach showed essentially the same 5-year survival by elective total gastrectomy (34.8%) as by partial gastrectomy (33.5%), while those with carcinoma of the distal 1/3 of the stomach showed a greater 5-year survival by elective total gastrectomy (43.7%) than by partial gastrectomy (29.8%). However, of significance is the fact that the incidence of nodal metastasis was 3 times greater in the patients undergoing elective total gastrectomy than in those undergoing partial gastrectomy. Despite this unfavorable finding, 5-year survival in the patients undergoing elective total gastrectomy for carcinoma of the mid 1/3 or distal 1/3 of the stomach was equal to, or better than, that found in those undergoing partial gastrectomy for lesions similarly located. On the basis of this finding alone, we believe that elective total gastrectomy is a worthwhile endeavor and should be performed for operable carcinomas arising in the mid 1/3 or distal 1/3 of the stomach.