Thyrotropin (thyroid-stimulating hormone or TSH)-receptor antibodies, important in the pathogenesis of Graves' disease, can be assayed by one of two methods: (1) bioassays that measure stimulation of thyroid cellular activity by patient immunoglobulins or (2) radioreceptor assays that measure inhibition of binding of labeled TSH to TSH receptors by the same substances. In this study, we report our experience with bioassay of thyroid-stimulating immunoglobulins (TSI) based on measurement of generation of cyclic adenosine monophosphate in a clone of the Fisher rat thyroid cell line (FRTL-5) in 279 patients, and we compare, in 163 consecutive samples, the results obtained by a radioreceptor assay for thyrotropin-binding inhibiting immunoglobulins (TBII). Among the untreated, hyperthyroid patients with Graves' disease, TSI were present in 95% (38 of 40), and TBII were present in 85% (17 of 20). In patients with euthyroid Graves' disease, TSI were found in 57% (16 of 28), and TBII were present in 41% (7 of 17). Of 49 nongoitrous and euthyroid controls, only 4% had TSI and 3% had TBII. Extremely high TSI indices were found in all patients who had pretibial dermopathy (N = 10) or severe Graves' ophthalmopathy (N = 19) requiring orbital decompression. We conclude that both assays are highly sensitive and specific in diagnosing Graves' disease. The TSI bioassay was more sensitive (P less than 0.001) than the TBII radioreceptor assay in detection of Graves' disease. In our experience, both assays have proved useful in the diagnosis of euthyroid Graves' disease with ophthalmopathy and atypical manifestations of hyperthyroid Graves' disease.