OBJECTIVE: Cord blood islet autoantibodies in children born to mothers with type 1 diabetes may be associated with a reduced risk of islet autoimmunity and diabetes. The aim of this study was to investigate in children with type 1 diabetes but born to non-diabetic mothers whether islet autoantibodies at birth affected their presence at diagnosis. PATIENTS AND METHODS: Serum samples at birth and at diagnosis were available from 141 children who developed type 1 diabetes between 1 and 19 yr of age (median 9.0 yr; male/female ratio 83/58). The samples were tested for autoantibodies against glutamic acid decarboxylase, insulinoma-associated antigen 2, and insulin as well as for islet cell antibodies. The human leukocyte antigen genotype was also determined. RESULTS: The frequency of islet autoantibodies in the umbilical cord blood was 11% compared with 91% at diagnosis. Children with fewer islet autoantibodies at diagnosis were more likely to have had autoantibodies at birth (p = 0.02). Autoantibodies present in cord blood at birth were observed in 25% (3/12) of children with no islet autoantibodies at diagnosis, in 17% (7/42) of children with one or two antibodies at diagnosis, and in only 5% (4/86) of children with more than two antibodies, demonstrating an inverse relationship between autoantibodies at birth and at diagnosis (test for trend, p < 0.001). CONCLUSIONS: Our preliminary data suggest that exposure to cord blood islet autoantibodies may influence the presence of islet autoantibodies at the time of diagnosis of type 1 diabetes and explain why some type 1 diabetes children are islet autoantibody negative at clinical diagnosis.