Although irinotecan 350 mg m À2 is a standard option for relapsed/refractory advanced colorectal cancer, there is some evidence that suggests that a higher dose may be more effective, with acceptable tolerability, following 5-fluorouracil (5-FU). This study assessed the optimal dosing strategy for irinotecan, along with treatment efficacy and safety. A total of 164 patients with metastatic colorectal cancer progressing after failure on 5-FU or raltitrexed received either 350 mg m À2 irinotecan (Group A; n ¼ 36) or 250, 350 or 500 mg m À2 , according to individual patient tolerance (Group B; n ¼ 62) or based on risk factor optimisation (Group C; n ¼ 66). There were no complete responses. There was a trend towards a higher overall response rate in Group B (13%) than in Groups A (8%) and C (9%). Tumour control growth rate was high in all three groups: 58% in group A, 60% in Group B and 50% in Group C. A total of 34% of patients in Group B and 9% in Group C were able to receive a dose of 500 mg m À2 . Median duration of response and time to progression were significantly longer in Groups A and B compared with Group C. No significant between-group differences for any adverse events were seen, although there was a small trend towards better tolerability in Group B. Individual dose escalation based on patient tolerance may allow more patients to receive a higher irinotecan dose without causing additional toxicity and can be an appropriate patient management strategy. 1999). Median overall survival rates of up to 10 months are achievable when irinotecan is used in relapsed/refractory colorectal cancer (Shimada et al, 1993; Rothenberg et al, 1996 Rothenberg et al, , 1999Pitot et al, 1997;Rougier et al, 1997;Van Cutsem et al, 1999). Two European phase III trials investigating the efficacy and safety of irinotecan, following 5-FU failure in advanced colorectal cancer, have demonstrated significant improvements in survival compared with best supportive care and 5-FU (Cunningham et al, 1998;Rougier et al, 1998). The main adverse events accompanying treatment with irinotecan in these trials were diarrhoea, neutropenia, fatigue, nausea and vomiting.Although 350 mg m À2 as an intravenous infusion every 3 weeks is the standard recommended dosage of irinotecan, pharmacokinetic parameters of irinotecan-lactone and the active metabolite SN-38-lactone vary between individuals (Xie et al, 2002). This may be attributed to differences in the levels of the enzymes that metabolise irinotecan, notably carboxylesterase for SN-38. Furthermore, the variable interindividual patient exposure to SN-38 has been identified as an important determinant of toxicity .At the same time, there is convincing evidence of a doseresponse relationship, and therefore a rationale for increasing doses when possible. In a phase I trial by Abigerges et al (1995), there were two recommended doses: 350 mg m À2 without highdose loperamide and 600 mg m À2 with high-dose loperamide. With the exception of one responder treated at 260 mg m À2 , all objective responses...