Our objective was to assess the antitumoral activity and toxicity of irinotecan (CPT-11) 60-min i.v. infusion every 2 weeks as second-line monotherapy of advanced colorectal cancer. Two doses were studied (250 and 200 mg/m) according to the risk of developing toxicity. Two groups of patients were studied: high-risk group (HR, 200 mg/m, n = 45; Karnofsky score 60-80% and/or the record of prior pelvic irradiation) and low-risk-group (LR, 250 mg/m, n = 51; Karnofsky score >80% and without prior pelvic irradiation). The mean number of cycles per patient was 7: 6.6 (HR group) and 8.3 (LR group). Median RDI was 0.96. The overall response rate was 8.9% [95% confidence interval (CI) 2.5-21.2%; HR group] and 15.7% (95% CI 7.0-28.5%; LR group), respectively. The LR group showed two complete responses and a higher percentage of stable disease (56.9 versus 33.3% in HR group). The median survival was 7.1 months (95% CI 5.2-8.9 months, HR group) and 11.7 months (95% CI 8.4-15.1 months, LR group). The median time to disease progression was 3.2 months (95% CI 1.0-5.4 months, HR group) and 5.3 months (95% CI 3.8-6.7 months, LR group). Both CPT-11 treatments were well tolerated. Grade 3/4 toxicity incidence was low, e.g. granulocytopenia (7% of patients in HR group and 9% in LR group) and delayed diarrhea (18% of patients in HR group and 14% in LR group). We conclude that the treatment of patients with the adjusted dose of CPT-11 according to prognostic factors for toxicity resulted in the improved toxicity profile, but showed poorer efficacy outcome. Therefore, the dose reduction in patients with low performance and treated with radiotherapy needs further investigation to provide some new insights on the benefit:risk ratio of such treatment.