A prospective randomized controlled trial of adjuvant chemotherapy for colorectal cancer was performed in 41 institutions in the Tokai district. In the study, postoperative administration of Mitomycin C (MMC) followed by carmofur (HCFU) for more than three months (arm B) was compared to MMC alone (arm A). Of 173 patients subjected to curative resection (arm A = 80, arm B = 93) for colorectal cancer, 148 including 69 from arm A and 79 from arm B were evaluable. Before clinical evaluation of the effectiveness the background factors were analyzed by the chi square test and U-statics. There were no differences between them. At four years after surgery Kaplan-Meire's life table showed that the disease free survival of arm B was superior to that of arm A, but this difference was not statistically significant (generalized Wilcoxon test: P less than 0.10). However, the survival in arm B was significantly prolonged as compared with that in arm A (P less than 0.05). Side effects such as leukopenia and gastrointestinal distress caused by administration of MMC and pollakisuria and gastrointestinal distress caused by administration of HCFU were observed in approximately 35% of the patients. They were all slight or moderate and reversible.