A cooperative study on surgical adjuvant immunochemotherapy for prevention of postoperative recurrence of gastric cancer was carried out by Kondo's group from July 1974 to December 1977. A total of 848 patients with gastric cancer underwent curative resection were eligible. Of 848 patients, 819 cases with complete description of patient's background were evaluable: Group A (surgery + MMC + 5-FU:chemotherapy) 253 cases, group B (surgery + MMC + 5-FU + OK-432 or PSK:immunochemotherapy) 282 and group C (surgery alone) 284. The 5-year corrected survival rates of total cases using Cox's proportional hazard model were 76.2% with group A, 73.6% with group B and 73.3% with group C. And the 10-year survival rates were 67.6% with group A, 64.3% with group B and 63.9% with group C. The difference in survival rate among groups A, B and C was not statistically significant. In the evaluation of prognostic factors by Cox's proportional hazard model, level of curative resection, extent of resection, histological stage and Kajitani's classification had an influence upon survival rates, in which degrees of metastasis in lymph node and serous infiltration were also involved. The significance of postoperative adjuvant immunochemotherapy has not been clarified.