Type 1 diabetes (T1D) is one of the most common chronic diseases of childhood and adolescence. The disease is considered to be of autoimmune origin. Its clinical manifestation represents end-stage insulitis, being a consequence of progressive destruction of the insulin-producing beta cells in the islets of Langerhans after an asymptomatic period of variable duration. T1D is a multifactorial disease with a strong genetic predisposition, the most important contributory genes being located in the HLA-DQ locus on the short arm of chromosome 6, 6p21. Autoantibodies can be detected in normoglycemic preclinical subjects and are used as markers of beta-cell autoimmunity. The present patient series was derived from the population-based Diabetes Prediction and Prevention Project (DIPP), which was initiated in 1994 to find optimal ways to identify children with an increased genetic risk of developing overt T1D from among the general population and to establish effective tools for preventing or delaying progression to the clinical disease. Genetic screening was performed on infants born at the university hospitals of Turku, Oulu and Tampere, and those with HLA DQB1-defined susceptibility to T1D were invited for an immunological follow-up. If persistent markers of beta-cell autoimmunity appeared, the family was offered the opportunity of enrolling the child for a randomized, double-blinded, controlled intervention trial with daily administration of intranasal insulin. HLA DQB1 genotypes were observed to be powerful determinants of beta-cell autoimmunity, affecting both the quantity and quality of autoimmunity. Autoantibodies appeared early in life and their frequency increased at least up to the age of 5 years. There was clear geographical variation in the cumulative incidence of autoantibodies among the regions studied here which could not be explained by differences in HLA DQB1genotypes or in the frequency of affected first-degree relatives. The appearance of autoantibodies also showed clear familial aggregation. ICA were observed to be a sensitive marker of beta-cell autoimmunity in our hands, showing good stability and identifying all those with a destructive beta-cell process, i.e. those who progressed to clinical diabetes. The number of progressors was limited, but the preliminary results indicated that IA-2A were the most specific autoantibody reactivity and also had the highest positive predictive value (PPV). The present experiences provide support for the relevance of grading diabetes risk in relation to HLA DQB1 genotypes, as this risk correlated well with the appearance of autoantibodies. Sequential monitoring of diabetes-associated autoantibodies makes it possible to identify high-risk individuals at a young age, facilitating the initiation of preventive trials and possibly preventive measures in the future, early in the disease process. The geographical difference in the appearance of autoantibodies might be explained by environmental factors. This population should be followed up further to assess whether the same difference can also be seen in the incidence of clinical disease and to evaluate whether possible environmental factors that either protect the children from the autoimmune process or predispose them to it might vary on a regional basis. A clear familial aggregation of autoantibodies was already seen in the present population of young children, and the families having at least two children with markers of beta-cell autoimmunity should be observed further in order to define the factors that distinguish non-progressors from progressors. ICA analyzed in a laboratory with a high standard of quality control showed high sensitivity in identifying children with destructive beta-cell autoimmunity and were stable enough, and therefore, in our experience, ICA are still the most sensitive of the single autoantibody reactivities for screening for beta-cell autoimmunity. IA-2A are highly specific autoantibodies, but they cannot be used as a screening tool alone due to their rather poor sensitivity.