Fibrosing colonopathy was first described in cystic fibrosis (CF) children in 1994. We have done a nested case-control study to identify possible associations with this condition. A case ascertainment within the UK CF population to identify any cases that occurred between January, 1984, and April, 1994, found 14 cases, all under 14 years and confirmed by independent histopathological review. All had presented since April, 1993; 12 were boys and six had received some or all of their care in Liverpool. Each case was matched, by date of birth, with four controls from the UK CF Registry. Information was obtained about cases and controls from their case records and by a structured interview with the families. In the 12 months before surgery, there was an association between the occurrence of fibrosing colonopathy and use of high-strength pancreatic enzyme preparations. This association was dose related. Odds ratio per extra 1000 high-strength capsules was 1.45 (95% CI 1.14-1.84). For use of protease, the odds ratio per million extra units per kg was 1.55 (1.19-2.03). For usage of individual high-strength products at any time during the 12 months before surgery some differences were observed; for Creon 25000 the odds ratio was 0.38 (0.10-1.42), for Nutrizym 22 43.4 (2.51-751), and for Pancrease HL 8.4 (1.95-36.1). These last two confidence intervals are extremely wide and compatible with these two products having the same odds ratios. Laxative use was independently predictive (odds ratio 2.42 [1.20-4.94]). We conclude that there is a dose-related association between high-strength pancreatic enzyme preparations and fibrosing colonopathy.