SUMMARY. By August 1997, 11,749 patients with cystic fibrosis had been enrolled in the European Epidemiologic Registry of Cystic Fibrosis (ERCF). Genotype analysis had been performed on 8,963 (76%) of these patients, and the majority had one or two identifiable mutations. Patients with known mutations were classified according to the type of mutation (Classes I-V), and were grouped according to the class of mutation on both chromosomes. This resulted in six subgroups, including all patients homozygous for Class I (I/I, n = 72), for Class II (II/II, n = 5,020), and for Class III mutations, (III/III, n = 23). Since there were only 23 patients homozygous for Class III mutations, a fourth group was made up of patients who were compound heterozygous for a Class II and III mutation (II/III, n = 265). There were only five patients homozygous for Class IV mutations, and consequently a fifth group was made up of all patients carrying at least one Class IV mutation, regardless of the nature of the mutation on the other chromosome (IV/any, n = 187). None were homozygous for Class V mutations; consequently, a sixth group consisted of patients carrying at least one Class V mutation (V/any, n = 22). Mean age was highest in groups III/III, IV/any, and V/any (15.6, 16, and 17 years, respectively) as opposed to 12.4 years in group II/II and 13.4 in group II/III, but both group III/III and V/any were small, and the confidence interval of the mean was large. The percentage of patients receiving pancreatic enzymes was lower in groups IV/any and V/any than in any of the other groups, i.e., approximately 50% of patients 18 years or older in both groups as opposed to between 90-100% of all other patients regardless of age. The prevalence of diabetes mellitus increased with age from 2.6% in patients < 18 years to 22.1% in patients 18 years or older in the large group II/II, but was only 1.5% in patients 18 years or older in group IV/any. Disregarding the small group III/III, abnormally elevated liver enzymes and/or bilirubin (1.5 x upper normal limit) was much less frequent in group IV/any than in any of the other groups, both overall and in patients aged 18 years or more. The course of lung disease appeared to be less dependent on genotype than pancreatic function, with only minor differences between groups; however, the mean values of both FVC % and FEV(1) % were slightly higher in group IV/any than all other groups in both younger and older patients. The same was found for the prevalence of some major clinical signs of severe lung disease, such as clubbing, hyperinflation, and crepitations. Overall mean weight expressed as an age percentile was markedly higher in group IV/any than in any other group, which may be related to the finding of a much lower prevalence of chronic P. aeruginosa infection in patients 18 years or older belonging to group IV/any (and V/any) than in any other group. In conclusion, the presence of a class IV mutation appears to offer some degree of protection against pancreatic insufficiency, diabetes mellitus, and liver disease. We confirmed that lung disease follows a milder clinical course in patients with a class IV mutation and that the presence of a class IV mutation (and possibly class V) is associated with a delay in the onset of P. aeruginosa infection., (Copyright 2001 Wiley-Liss, Inc.)