The aetiology of perforation of large bowel diverticula is poorly understood and a case-control study is required to identify the causes. Before such a study can be attempted, the incidence must be determined and groups at particular risk identified. Cases of perforated large bowel diverticula living in the Norwich postal code region treated between 1995 and 1997 were identified. Fifty-eight cases presented in a population of 531 241. The incidence was 4.0 cases per 100,000 per year, increased with age and was higher in men than women (5.8 vs 3.1). The most frequently used drugs were non-steroidal anti-inflammatory drugs (NSAIDs) (29%) and opiate analgesics (26% of cases). This is the first report of the incidence of perforated diverticular disease and allows a calculation of the population size needed to recruit sufficient cases for an aetiological investigation. The differences in incidence between genders should prompt a search for factors which differ between the sexes such as diet. NSAIDs are a known risk factor, although the data show that opiate analgesics should be investigated.