Background: Results on overall survival in randomised studies of mono- versus combination chemotherapy in advanced colorectal cancer patients may have been biased by an imbalance in salvage treatments. This is the first randomised study that evaluates sequential versus combination chemotherapy with a fluoropyrimidine, irinotecan and oxaliplatin. Patients and methods: A total of 820 patients were randomised between first-line capecitabine, second-line irinotecan and third-line capecitabine + oxaliplatin (arm A) versus first-line capecitabine + irinotecan, and second-line capecitabine + oxaliplatin (arm B). The primary end point was overall survival. We present the results of an interim analysis on the safety data in the first 400 patients. Results: In first-line the incidence of grade 3–4 diarrhoea, nausea, vomiting and febrile neutropenia was significantly higher in arm B. However, when toxicity over all lines was considered only grade 3 hand–foot syndrome occurred more frequently in arm A (12% versus 6%, respectively, P = 0.041). The incidence of cardiovascular toxicity was low. In two out of five patients with sudden death (one in arm A, four in arm B) cardiovascular risk factors were present. Conclusions: Both treatment arms had an acceptable safety profile. These data imply that the results on survival will be the major determinant for the selection of either strategy. Capecitabine plus irinotecan appears to be a feasible first-line treatment for patients with advanced colorectal carcinoma.