A prospective, randomized, controlled study of adjuvant chemotherapy in operable gastric cancer was commenced in 1976. Four hundred and eleven patients have been randomized into 1 of 3 treatment groups. Group A received a placebo injection of intravenous normal saline (10 ml) at 3-week intervals. Group B received a 5-day induction course of 5-fluorouracil (5-FU), vincristine, cyclophosphamide, and methotrexate followed by 3-week intravenous injections of 5-FU and mitomycin C (MMC). Group C received 3-week intravenous injections of 5-FU and MMC. The treatment was to be given for 2 years. Patients were stratified for age, sex, preoperative duration of symptoms, and clinicopathologicai stage prior to randomization. At this interim analysis, preoperative duration of symptoms and clinicopathological stage were statistically significant prognostic factors (p = <0.05 and <0.001, respectively).At the time of this analysis, there has been no significant improvement in survival in either of the treatment groups compared to the control group. While treatment could begin at any time during the first 3 postoperative months, survival at 1 year was significantly improved (p = 0.013) by treatment C in those who started treatment within I month of surgery; however, this benefit of treatment was not maintained at 3 years. Of the 279 deaths, 16 (5.7%) have been drug-related, 11 resulting from renal toxicity associated with this therapeutic regimen.This study has confirmed the poor prognosis of gastric cancer, and the early results demonstrate that there is no m