The primary objective was to compare the 3-year survival of rectal cancer patients randomised postoperatively to irinotecan (IRI), Leucovorin (LV) and bolus 5-fluorouracil (5FU) or LV-bolus 5FU with radiotherapy. Secondary objectives included disease-free survival, local relapse and toxicity. The study included 321 eligible patients. The treatment consisted of weekly administration of IRI 80 mg/m(2) intravenously (IV), LV 200 mg/m(2) and 5FU 450 mg/m(2) bolus (arm A) versus LV 200 mg/m(2) and 5FU 450 mg/m(2) IV bolus (arm B). One cycle included four infusions and treatment was continued for a total of six cycles. The first cycle was followed by pelvic irradiation plus 5FU. There were no differences between the arms in 3-year overall, disease-free and local relapse-free survival. Grades 3 and 4 toxicity was similar in both the arms with the exception of leucopaenia, neutropaenia and alopecia, which were higher in the IRI arm. IRI added to adjuvant radiochemotherapy with LV and bolus 5FU was not shown to improve survival, whereas the incidence of severe leucopaenia was significantly higher in the IRI arm.