A two-site immunoradiometric assay for serum thyrotropin (TSH) was modified to improve the analytical sensitivity. The sensitivity achieved (detection limit, approximately 0.1 microU/ml; lower limit of quantitative measurement, approximately 0.4 microU/ml) was comparable to that of the best competitive binding research assays, yet this assay can be performed routinely. Serum TSH was 1.82 +/- 0.69 (mean +/- s.d.) (range 0.4-3.4 microU/ml) in healthy individuals and 1.83 +/- 0.90 microU/ml (range 0.7-3.7 microU/ml) in patients with nonthyroidal disorders. By contrast, 97% of clinically hyperthyroid patients (Graves' disease, toxic nodular goiter) with high serum free T4 (FT4) and T3 had suppressed serum TSH values, i.e., less than 0.3 microU/ml. Among patients with euthyroid Graves' ophthalmopathy or nontoxic goiter those clinically suspected of mild hyperthyroidism had TSH values less than 0.3 microU/ml, while those judged euthyroid had normal values. A large proportion of thyroid patients on antithyroid drugs (poorly to well-controlled) had suppressed TSH. Of Graves' patients in remission (normal FT4 and T3), 75% had normal TSH, but individual levels changed significantly over time, suggesting that a progressive decline in TSH may be useful in predicting recurrences. In hypothyroid patients taking L-T4, serum TSH was subnormal in patients with elevated FT4, but TSH was also low in six patients clinically suspected to be thyrotoxic despite normal FT4 and T3 and in 32% of asymptomatic patients with normal thyroid hormone levels. Conversely, 23% of thyroid cancer patients who had undergone thyroidectomy were taking insufficient L-T4 to completely suppress TSH secretion. In 25 individuals who underwent thyrotropin releasing hormone (TRH) stimulation tests, the baseline serum TSH value correlated well with the peak serum TSH value post-TRH (r = 0.85). We conclude that sensitive TSH measurements could establish or confirm the diagnosis of hyperthyroidism in equivocal cases, replace most TRH-stimulation tests and be of value in optimizing L-T4 suppression therapy for thyroid cancer patients post-thyroidectomy.