Abstract. It has been proposed elsewhere that antithyroid drugs (ATD) bring about remissions in Graves' disease (GD) by directly suppressing the immune system. The decline in thyroid stimulating antibody (TSAb) and other thyroid autoantibodies, and the reduction in intra-thyroidal lymphocytes has been attributed to this effect, which has been considered to be via interference with oxidative reactions within the antigen-presenting cell. However, if ATD exert an immunosuppressive effect, one might expect that this would be a pharmacological action, paralleling the overall pharmacological effect of normalizing thyroid function. Moreover, there is evidence that in Hashimoto's thyroiditis (HT) thyroid autoantibodies do not decline as a result of such treatment. We propose herein that ATD act through modulation of thyroid cell activity, which in turn has an effect on the immune system so as to induce immunological remissions in GD. It has been shown that hyperthyroidism itself has a deleterious effect on generalized suppressor T (Ts) lymphocyte function which is additive to the organ-specific defect in Ts function. Normalization of thyroid function tends to normalize non-specific Ts function; suppression of the thyroid-specific helper T cell (Th) population is thus brought about, in that subset of patients who do not have a severe organ specific Ts defect. This results in a reduction of interferon γ production, as well as reduced Th stimulation of B lymphocytes to produce TSAb. The combined reduction of TSAb and interferon γ will then reduce thyroid antigen and HLA-DR expression, i.e. antigen presentation by the thyroid cell, as well as a further reduction in thyroid hormone production. The latter will further reduce Ts inhibition, reducing Th activity; reduced antigen presentation will also reduce Th activity. This beneficial cycle will then lead to remissions, except in those patients with a more severe organ specific Ts defect.