Autoantibodies of immunoglobulin G (IgG) class to the thyrotropin receptor (TSHR) are a paradigm for humoral, organ-specific, autoimmune disease. Unlike diseases such as diabetes mellitus type I, in which the role of antibodies to a variety of antigens is not clearly defined, and in which cellmediated immunity appears to be dominant, the direct pathogenetic relationship between TSHR autoantibodies and human disease has been clearly evident for more than two decades (reviewed in 1). Evidence for light chain restriction among TSHR autoantibodies in an individual serum suggests that, at least in some patients, these antibodies are oligoclonal (2-4). More recently the molecular cloning of the TSHR has led to the availability of large quantities of pure antigen of bacterial or insect cell origin, or as synthetic peptides. At face value, therefore, characterization of the TSHR autoantibody repertoire in human disease should be straightforward. In reality, however, TSHR autoantibodies have proven extraordinarily difficult to study. One major reason for this difficulty is the very low TSHR autoantibody titer in serum. This great difference in TSHR and thyroid peroxidase (TPO) autoantibody concentrations in serum is not generally recognized. Only rare sera of exceptionally high titer are capable of “lighting up” the TSHR by indirect immunofluorescence on cultured thyroid cells (5), consistent with our own inability to use Graves’ sera to immunoprecipitate the TSHR on thyroid cells (6) or to detect the TSHR on transfected nonthyroidal cells by immunofluorescence (unpublished data, B. Rapoport). We have suggested that the low TSHR autoantibody titer, as well as autoantibody oligoclonality, reflects the clinical presentation of Graves’ hyperthyroidism at an early phase of the autoimmune response (7). Thus in Graves’ disease, the TSHR is an Achilles’ heel for clinical symptoms, whereas hypothyroidism may only manifest at a later stage in Hashimoto’s thyroiditis when thyroid destruction ultimately overwhelms the regenerative capacity of the gland. Another difficulty contributing to TSHR autoantibody complexity is heterogeneity of the antibody population. There are thyroid stimulatory autoantibodies and (more rarely) TSH blocking antibodies (reviewed in 1, 8). As would be anticipated, the regions on the TSHR recognized by these different autoantibodies are not identical (9, 10). However, the situation is more complex. Even among stimulatory TSHR autoantibodies there is epitope heterogeneity(11-13). Further, as suggested previously (7), TSHR autoantibodies may also exist that neither stimulate the receptor nor inhibit