366 patients fully resected from a Dukes B2 or C colorectal cancer were randomised to receive 6 courses of systemic chemotherapy comprising either 5-fluorouracil (5 FU) alone (arm A: 450 mg/m2/day-5/21 days) or combined folinic acid (FOL) and 5 FU (arm B: respectively 200 mg/m2 racemic form or 100 mg/m2-l-form and 370 mg/m2/day-5/21 days). 173 patients had also been initially randomised to receive one course of intraportal chemotherapy just after surgery or no portal treatment. Oral levamisole (150 mg/day; 3 days every other week) was given to all patients for one year. A significantly higher incidence of leuco-granulocytopenia was observed in the arm A (5 FU alone) inducing more frequent dose delays and adaptations as well as levamisole's withdrawal. Then dose-intensities and dose-intensity products were lower in this arm but the dose intensity expressed in mg/m2/week remained higher (631 +/- 107 vs 557 +/- 99; p < 0.001). The median follow-up in the study was 4.5 years. Relapse free (RFS) and overall survivals (OAS) were prolonged in the 5 FU alone group peculiarly in those patients who had not been randomised for portal treatment. Curves diverged progressively with longer follows-up (at 8 years; RFS in arm A: 67-71% vs 59-53% in arm B; OAS in arm A: 72-74% vs 56-46% in arm B). Patients suffering from a colon or a Dukes C cancer benefited the most from the treatment with 5 FU alone. The results are discussed in the light of other recent adjuvant trials. Well dosed 5 FU over a short period of time without folinic acid may be a valuable and inexpensive adjuvant treatment for colorectal cancer. Levamisole may no longer be recommended in this setting.