The autoimmune nature of the diabetogenic process and the major contribution of T lymphocytes stand now beyond any doubt. However, despite the identification of the three major type 1-diabetes-related autoantigens (insulin, GAD65 and phosphatase IA-2), the origin of this immune dysregulation still remains unknown. More and more evidence supports a thymic dysfunction in the establishment of central self-tolerance to the insulin family as a crucial factor in the development of the autoimmune response selective of pancreatic insulin-secreting islet beta cells. All the genes of the insulin family (INS, IGF1 and IGF2) are expressed in the thymus network. However, IGF-2 is the dominant member of this family first encountered by T cells in the thymus, and only IGFs control early T-cell differentiation. IGF2 transcription is defective in the thymus in one animal model of type 1 diabetes, the Bio-Breeding (BB) rat. The sequence B9-23, one dominant autoantigen of insulin, and the homologous sequence B11-25 derived from IGF-2 exibit the same affinity and fully compete for binding to DQ8, one class-II major histocompatibility complex (MHC-II) conferring major genetic susceptibility to type 1 diabetes. Compared to insulin B9-23, the presentation of IGF-2 B11-25 to peripheral mononuclear cells (PBMCs) isolated from type 1 diabetic DQ8+ adolescents elicits a regulatory/tolerogenic cytokine profile (*IL-10, *IL-10/IFN-g, *IL-4). Thus, administration of IGF-2 derived self-antigen(s) might constitute a novel form of vaccine/immunotherapy combining both an antagonism for the site of presentation of a susceptible MHC allele, as well as a downstream tolerogenic/regulatory immune response.