There is no room for complacency concerning our knowledge of Crohn's disease of the colon, and this applies to epidemiology as to other aspects of diagnosis and management. In this presentation therefore "current knowledge" is first reviewed. Crohn's disease of the colon is shown to be a worldwide disorder with incidence rates of between one and four cases per 100,000 population per year and prevalence rates up to ten times this figure. This disease is most commonly found in Europe and North America, with a slight excess in females over males and a preponderance of young persons whose disease begins in early adult life. Two-thirds of cases have colonic involvement. Racial and ethnic studies have been confusing and there is some reason to doubt the simplistic assertion of the 1960's that the disease is most common in Jews. Genetic marker studies have been similarly inconclusive despite an apparent association with HLA-B27-as have most studies involving environmental factors such as diet, smoking and the contraceptive pill. The presentation next explores some important reasons why this current "knowledge" is suspect as well as inconclusive. There is a total lack of etiological data about Crohn's disease, and until recently no standard definition of the disease existed. Many epidemiological studies have selected cases from small groups rather than the whole population-a particularly difficult problem due to the protean nature of the disease. Observer variation in eliciting data presents a further problem as does the simplistic nature of the analysis carried out to date. Finally, some tentative solutions are proposed for current outstanding problems. It is argued that Popperian philosophy may form the basis for a working description of Crohn's disease and the World Organisation of Gastroenterology Crohn's disease study has proposed such a definition. Multi-national trials are clearly necessary if some of the problems listed earlier are to be overcome; as is the development of reproducible and relevant indices for assessing the severity of disease. However, in the last analysis the only real key to progress is an improvement in our fundamental understanding of the disease itself; and at the present time there seems very little prospect of progress on this front. Until such progress is made, the epidemiology of Crohn's disease will remain a difficult and muddled area.