TSH secretion is decreased by both T4 and T3. This negative feedback control of TSH secretion has been correlated with an increase in pituitary nuclear T3 content, and it is not clear whether T4 exerts its effect directly on the thyrotroph or after its deiodination to T3. However, levels of the pituitary enzyme catalyzing T4 to T3 conversion, 51)-II, are decreased in the presence of an increased amount of T4. Thus, it is unclear why the thyrotroph would have a mechanism for modulating the production of T3, if T3 is, in fact, the sole bioactive signal providing negative feedback inhibition.To examine this apparent paradox, we administered EMD 21388, a compound which inhibits the binding of T4 to transthyretin resulting in a rapid increase in circulating free T4 levels, to rats pretreated with radiolabeled T4 and T3. We observed increases in pituitary and liver T4 content of >150%, without increases in the respective tissue T3 contents. The EMD 21388-treated rats also exhibited a 25% decrease in pituitary 51)-II activity (103.8±15.8 fmol 125I released * mg protein-' h-1, vs. control, 137.4±15.9, mean±SE), as did rats treated with sodium salicylate, another compound that inhibits T4-TTR binding (100.8±7.1). TSH levels significantly decreased 2 h after the administration ofEMD 21388. These data demonstrate that despite a T4-mediated decrease in pituitary 5D-II activity, an increase in 14 independently decreases TSH secretion. (J. Clin. Invest. 1991. 88:898-903.)