The effect of graded increments of chronically administered oral T4 or T3 on the TSH response to TRH was studied in normal young adult men. The TSH response was assessed in the baseline state and after each increment of each hormone (two weeks at each dose level) using both 30 mug and 500 mug doses of TRH. Each thyroid hormone caused a dose-related decrease in the TSH response to TRH; thus the TSH response could be used as a bioassay for the biologic activity of the thyroid hormones in man. The dose of thyroid hormone that caused a 50% suppression of the TSH response, or the SD50, was not different with either 30 mug or 500 mug of TRH indicating that thyroid hormone suppression of the TSH response is not more easily detected with a small dose of TRH. The mean SD50 for T4 was 115 mug/day, for T3 stopped 2 h before testing the mean SD50 was 29 mug/day, and for T3 stopped 24 h before testing it was 45 mug/day. Using the average SD50 for the two T3 regimens (37 mug/day), the calculated relative potency indicates that oral T3 is 3.3 times as potent as oral T4, a value in reasonable agreement with the value previously estimated with a calorigenic end-point. The mean dose of T4 needed to decrease the TSH response to TRH to below the normal range (max delta TSH of 2 muU/ml) was 150 mug/day; this value is probably more appropriate than the SD50 in the treatment of patients with primary hypothyroidism or goiter and was about the same (160 mug/day) using a peak TSH after TRH of 3 muU/ml as an end-point. Estimation of the SD50 in each subject showed a 2- to 3-fold range with all regimens of thyroid hormones; similarly there was a 2-fold in the dose of T4 needed to suppress the TSH response to TRH to below the normal range. Further, the difference in the mean SD50 for the two T3 regimens indicates that a single daily dose of oral T3 does not exert a constant biologic effect throughout the day. Thus, because of individual variation and, in the case of T3, because of changing activity during the day a given dose of thyroid hormone may have a widely varying biologic effect. There was also a 3-fold range in the relative potency of T3 to T4 in the four subjects treated with both hormones. This suggests that the therapeutic administration of a fixed ratio fo T3 to T4 may have a variable effect from patient to patient. Finally, the serum T4 rose while the serum T3 did not at a dose of T4 that abolished the TSH response to TRH, indicating that circulating T4 is a determinant of TSH secretion in normal man.