Free thyroxine (FT4) and free triiodothyronine (FT3) concentrations in serum were measured by direct equilibrium dialysis methods in patients receiving thyroxine replacement or suppression therapy. Four of 50 hypothyroid patients euthyroid on replacement therapy (mean thyroxine dose 120 micrograms/day) had supranormal FT4 concentrations, whereas the FT3 concentrations were normal in all. Forty-one of 56 operated thyroid carcinoma patients on suppressive therapy (mean thyroxine dose 214 micrograms/day) had raised FT4 concentrations, whereas the FT3 concentrations was elevated in only one patient. There was a large difference in mean FT4 values for hypothyroid and thyroid carcinoma patients (17.2 vs 29.5 pmol/l), whereas the difference in mean FT3 values was small (5.0 vs 6.1 pmol/l), suggesting a decreased peripheral conversion of T4 to T3 with increasing concentrations of FT4. Serum TSH concentrations, as determined by an immunoradiometric assay, varied from less than 0.02 to 11.9 mU/l in treated hypothyroid patients; 21 patients (42%) had values outside the reference limits. As a single test, serum TSH is therefore not very useful for the assessment of adequate thyroxine dosage in patients with primary hypothyroidism. In thyroid carcinoma patients, the TSH concentrations were less than 0.18 mU/l; 45 patients had values less than 0.02 mU/l indicating sufficient suppression of TSH secretion in the majority of cases. On the basis of these results we recommend the combination of FT3 and TSH tests for monitoring thyroxine replacement and suppression therapy. FT4 appears less useful than FT3 for this purpose even if special reference values values were adopted for each patient group.