We have hypothesized over many years that Graves' disease (GD) and the other autoimmune thyroid diseases (AITD) are each due to antigen-specific defects in suppressor (regulatory) T lymphocyte function. There have been several reports dealing with the role of regulatory T lymphocyte subsets, ie., that will prevent autoimmune disease in these and other organ-specific autoimmune diseases. In AITD, suppressor T cells have been shown to be less well activated by relevant antigen, but are normally activated by irrelevant antigen; suppressor T cells from normal persons react equally well to both. In GD, these cells have been shown to be inadequately activated by TSH receptor antigen, but are normally activated by irrelevant autoantigen. This reduction is partial only, and insufficient itself to precipitate the autoimmune disease; further insults from the environment are necessary to further reduce generalized regulatory cell activity, adding to the genetically induced specific regulatory cell dysfunction, which appears in turn to be due to a specific defect in the presentation of a specific antigen. This, in turn, may relate to abnormalities of the genes responsible for antigen presentation. The end result is activation of appropriate helper and effector T cells, the stimulation by these of appropriate B lymphocytes, and the concurrent production of cytokines. These events lead to functional changes within the target cell which itself will express Class II antigens, heat shock proteins, and intercellular adhesion molecules, all of which amplify the immune response. Moreover, the activation of helper T lymphocytes by specific antigen depends on the availability of normal amounts of antigen being presented to them by antigen-presenting cells. Thus, there is no need to invoke any primary abnormality or infection of the thyroid cell, or any cross-reacting antigen of microorganismic origin to initiate this process. What is required is an abnormality of antigen-presentation such that regulatory cells are not properly activated, plus some additive environmental disturbance acting on the immune system. GD specifically results from the production by B lymphocytes of an antibody directed against the TSH receptor which stimulates the thyrocyte in a manner similar to TSH, but for a much longer interval. There are also antibodies to the thyrotrophin (TSH) receptor which block the action of TSH. Thyroid stimulating antibody is typical of GD and is detectable in about 95% of cases, but is also seen in destructive thyroiditis transiently. It tends to decline with antithyoid drug therapy, and rises further (for several months) after 131 I treatment. It may slowly decline after subtotal thyroidectomy. It also declines in the third trimester of pregnancy but sometimes is sufficiently high to cause foetal and neonatal passive transfer GD. It tends to rebound in the mother after delivery and may result in postpartum GD. The blocking antibody may cause atrophic thyroiditis and hypothyroidism. Antimicrosomal antibody has now been shown to be antithyroperoxidase. It correlates moderately well with thyroid dysfunction in Hashimoto's thyroiditis (HT) and GD, while antithyroglobulin is of much less value. Graves' ophthalmopathy is still not well understood, and its precise relationship to Graves' hyperthyroidism has yet to be worked out. However the retroorbital fibroblast is now emerging as the most likely target cell, with retroorbital muscle involvement possibly secondary. A recent observation of a genomic point mutation on the TSH receptor on fibroblasts from patients with Graves' ophthalmopathy but not normal persons raises interesting possibilities.