There is excellent presumptive evidence that an IgG plays an aetiological role in the development of hyperthyroidism in Graves' disease but available methods for detecting thyroid stimulating immunoglobulins (TSI) are still far from satisfactory. They show considerable variation in specificity, sensitivity and precision, and comparison of the experimental data, obtained with these various methods, is difficult. A need exists for an in vitro TSI assay which is based upon the propensity of IgG molecules to stimulate the thyroid gland. Because a complex chain of biochemical events is involved, including binding to IgG to the cell membrane receptor, release of cyclic AMP, organification of iodide, hydrolysis of iodoproteins and secretion of thyroid hormones, it is not yet clear which step should be monitored to obtain the best index of thyroid stimulating activity. Although TSI and related assays appear to be of limited value in the primary diagnosis of Graves' disease, they offer some assistance to the clinician in the following situations: 1. Prediction of relapse in Graves' disease patients who have been rendered euthyroid with antithyroid drugs. 2. Identification of patients with ophthalmic Graves' disease. 3. Prediction of neonatal hyperthyroidism in thyrotoxic pregnancies.