In this report we summarize evidence to support a model for the development of Graves' disease. The model suggests that Graves' disease is initiated by an insult to the thyrocyte in an individual with a normal immune system. The insult, infectious or otherwise, causes double strand DNA or RNA to enter the cytoplasm of the cell. This causes abnormal expression of major histocompatibility (MHC) class I as a dominant feature, but also aberrant expression of MHC class II, as well as changes in genes or gene products needed for the thyrocyte to become an antigen presenting cell (APC). These include increased expression of proteasome processing proteins (LMP2), transporters of antigen peptides (TAP), invariant chain (Ii), HLA-DM, and the co-stimulatory molecule, B7, as well as STAT and NF-kappaB activation. A critical factor in these changes is the loss of normal negative regulation of MHC class I, class II, and thyrotropin receptor (TSHR) gene expression, which is necessary to maintain self-tolerance during the normal changes in gene expression involved in hormonally-increased growth and function of the cell. Self-tolerance to the TSHR is maintained in normals because there is a population of CD8- cells which normally suppresses a population of CD4+ cells that can interact with the TSHR if thyrocytes become APCs. This is a host self-defense mechanism that we hypothesize leads to autoimmune disease in persons, for example, with a specific viral infection, a genetic predisposition, or even, possibly, a TSHR polymorphism. The model is suggested to be important to explain the development of other autoimmune diseases including systemic lupus or diabetes.