Following curative resection for gastric adenocarcinoma, 142 patients were randomized to receive adjuvant chemotherapy of methyl-CCNU and 5-FU or to be closely followed without chemotherapy. The median patient has been followed for four years. At the present time, there have been 38 recurrences and 40 deaths among the 71 patients in the control arm and 28 recurrences with 29 deaths in the group of patients randomized to chemotherapy. Median survival for control patients is 33 months whereas it is estimated to be in excess of four years for the adjuvant therapy group. A survival advantage appears to be associated with adjuvant treatment, a test for treatment difference being significant (P less than 0.03) after adjustment for important covariates. Survival and disease-free intervals are longer in the chemotherapy arm for patients in all prospective stratification subgroups (proximal and total vs distal subtotal gastrectomy, presence or absence of lymph node metastases, invasion of adjacent organs vs confined to stomach wall and location of tumor within the stomach). A final assessment of this surgical adjuvant approach must await complete results of this trial, as well as of others currently in progress evaluating this same concept.