As part of a District-based audit of diabetes care all attendances (of diabetic and non-diabetic patients) at a local Accident and Emergency department were monitored over a 3-month period. A total of 9505 attendances took place of which 62 (0.7%) were by people with diabetes. This was no different from the expected number of attendances (61.5) based on the rates seen in non-diabetic individuals. Of these 62 attendances, 20 (by 15 individuals) were directly related to diabetes (hypoglycaemia 17, hyperglycaemia 1, for supplies of insulin or equipment 2). When these diabetes-specific attendances were removed from the total, the number of attendances (42) was significantly (p = 0.004) less than that expected. Diabetic males aged 0-19 years attended significantly more often than non-diabetic individuals of the same age but this excess was solely attributable to diabetes.