TSH binding-inhibiting antibody (TBIAb) activity was measured in 809 European patients with different forms of hyperthyroidism. Distribution of these TBIAb was skewed, with a peak in the range of normal controls, and an ill defined, not clearly separated peak at higher levels of TSH displacement. There was no unequivocal separation of two possible subgroups of hyperthyroidism (immunogenic and non-immunogenic). TBIAb distributions of patients with and without endocrine ophthalmopathy (EO) overlapped considerably. Although patients with Graves' disease, arbitrarily defined by the presence of endocrine ophthalmopathy or diffuse nuclide uptake by thyroid scanning, had mostly elevated TBIAb activity, 24.3% had values within the range of normal controls (mean + 2SD). Patients with diffuse thyroid uptake had significantly higher TBIAb levels than patients with nodular scan findings. In Graves' disease, TBIAb activity was positively correlated with the severity of endocrine ophthalmopathy, the size of the thyroid, and the serum levels of total and free triiodothyronine. There was no influence of age, sex, pretreatment, or regional iodine supply. These results suggest (1) that the clinical manifestations of Graves' disease are statistically related to TBIAb activity and (2) that separation of immunogenic and non-immunogenic forms of hyperthyroidism by means of TBIAb determination is unsatisfactory. The almost continuous distribution of TBIAb points to insufficient sensitivity of the present technique and raises doubts as to whether TBIAb values can be reliably classified as 'positive' or 'negative'.