Fifty patients with severe complications of diverticular disease were compared with two groups of 50 controls, matched for age and sex. The first control group (A) was randomly selected from all emergency hospital admissions, and the second group (B) from patients with uncomplicated diverticular disease. Of the 50 study patients, 24 (48 per cent) were taking non-steroidal anti-inflammatory drugs (NSAIDs) at the time of admission compared with nine (18 per cent) of control group A and ten (20 per cent) of control group B. Both of these differences were statistically significant, indicating a strong association between the ingestion of NSAIDs and the development of severe complications of diverticular disease.