If environmental factors are important in the aetiology of insulin-dependent (Type 1) diabetes mellitus, primary schooling would be more likely to be shared by cases in an epidemic year compared with controls. We have examined a case control study, comparing primary schools attended. Cases were identified from an established register with known accurate ascertainment and compared with controls taken from a central register of all children and date of birth, sex, and race matched. Parents were asked to complete a structured questionnaire which included questions relating to the child's schooling. There were 187 cases with at least a first matched control. Thirty-two of the cases attended the same nursery or primary school as another case, 18 in the same academic year, compared with 12 controls attending the same school as another control (99% confidence interval (CI) for the difference between proportions 0.0572 - 0.107). Of these cases diagnosed in 1986, 9 attended the same school as another case. Spatial clustering occurs in Type 1 diabetes. This is further evidence of an environmental factor contributing to the genesis of the disease.