Type 1 diabetes (T1D) is a chronic disease caused by destruction of the insulin producing beta-cells in the pancreas. The incidence of T1D has increased rapidly, especially in the Western world and among young children. The pathogenesis of T1D is not fully understood, but the beta-cells are believed to be destroyed by an autoimmune process initiated years before the onset of T1D. During this pre-clinical period, autoantibodies to insulin (IAA), glutamic acid decarboxylase (GADA) and the tyrosine phosphatase-like protein IA-2 (IA-2A) can be detected and are used to identify individuals at risk of T1D. The major genetic determinant for T1D is the HLA class II genes, but also polymorphism in the insulin gene and CTLA-4 gene are associated with T1D. The risk genes cannot explain the rapid increase in incidence of T1D, therefore a role for different environmental factors has been suggested. The aim was to study the prevalence of beta-cell autoantibodies in children from the general population in relation to known genetic and environmental risk factors, and in young patients with T1D in high and low incidence areas. Short duration of breast-feeding was associated with an increased risk of developing beta-cell autoantibodies in children from the general population at 5-6 years of age. We found an association between positivity for GADA and/or IAA at the age of 5-6 years and a short duration of total breastfeeding, and also between positivity for GADA, IA-2A and/or IAA and a short duration of exclusive breast-feeding. Our findings suggest that breast-feeding has a long term protective effect on the risk of beta-cell autoimmunity in children from the general population. The T1D related risk genes were not associated with beta-cell autoantibodies other than GADA in children from the general population at 5-6 years of age. Children with the DR4-DQ8 haplotype were more often positive for GADA than children without this haplotype. We found no association of GADA with DR3-DQ2 haplotype or between these two haplotypes and any of the other autoantibodies. Our results suggest that beta-cell autoimmunity in children from the general population is not strongly associated with any risk genes of T1D other than DR4-DQ8. In the non-diabetic children with allergic heredity GADA was detectable in almost all children, IA-2A in about half and IAA in 10% of the children. The levels low of these autoantibodies fluctuated with age and different patterns of fluctuations were seen for GADA and IA-2A, which may reflect differences in the immune response to the autoantigens. In patients with newly diagnosed T1D, we found some differences between patients from a high incidence country (Sweden) and a country with a lower incidence (Lithuania). Among the Swedish patients, the prevalence of IAA and GADA or multiple autoantibodies was higher than in Lithuanian patients. The risk genes DR4-DQ8 and the heterozygous high risk combination DR4-DQ8/DR3-DQ2 was more common among the Swedish patients than Lithuanian patients. Patients with low levels of IAA had higher levels of HbA1c and ketones, indicating that patients without IAA or with low levels of IAA have a more severe onset of T1D. Our findings indicate that beta-cell autoimmunity is more pronounced in a high incidence area compared to an area with a lower incidence. In conclusion, short duration of breast-feeding is a risk factor for beta-cell autoantibodies in children from the general population, and the beta-cell autoantibodies in these children are not associated with specific risk genes. Children with newly diagnosed T1D in a high incidence area carry risk genes and have autoantibodies more often than newly diagnosed children from an area with a lower incidence, perhaps indicating different disease phenotypes.