Abundant evidence now exists that autoimmunity plays a critical role in the pathogenesis of type 1 (insulin-dependent) diabetes mellitus. The non-obese diabetic (NOD) mouse is an extensively studied animal model of this T-cell-mediated autoimmune disease. Our laboratory has focused on isolating diabetogenic T cell clones from NOD mice as a means of elucidating the pathogenesis of type 1 diabetes. This experimental approach presupposes that type 1 diabetes in NOD mice results from the action of islet-reactive T cells that are not present in other mouse strains; the diabetogenic T cells would therefore represent "forbidden clones" which exist in NOD mice as a result of a failure of clonal deletion. While the inappropriate presence of diabetogenic T cells probably plays a central role in murine diabetes, it cannot explain all aspects of the disease. Type 1 diabetes is a chronic disorder with a lengthy preclinical stage; if the diabetogenic T cells acted in an unopposed fashion, one might expect to see a much more fulminant clinical course. This observation suggests that regulatory influences are likely to exist in this disease--a possibility supported by recent experimental data. If these regulatory influences could be identified and enhanced, specific immunotherapy for type 1 diabetes could be achieved.