A case-control study was conducted to study the proposed inverse relationship between breast feeding and incidence of Type 1 diabetes mellitus. All Danish diabetic men born in Copenhagen during 1959-1964 and/or residing there for the first year of life (n = 119) were identified and a search made for their post-natal health visitor records. These include data on birth weight, birth length, immunizations, and feeding habits. A total of 77 cases were re-identified. No differences were found with respect to clinical characteristics between these cases and those not identified. For each diabetic subject, two control children with date of birth identical to the diabetic cases were drawn from health visitor records. Cases and controls did not differ with respect to maternal age, or birth weight or length. Overall, the statistical analysis failed to confirm the hypothesis of an association between duration of breast feeding and subsequent diabetes risk.