Studies of the possibility that thyroxine (T4) is converted to 3.5,3'-triiodo-L-thyronine (T3) in the extrathyroidal tissues in man have been conducted in 13 patients, all but two of whom were athyreotic or hypothyroid, and all of whom were receiving at least physiological replacement doses of synthetic sodium-L-thyroxine.T3 was found in the sera of all patients, in concentrations ranging between 243 and 680 ng/100 ml (normal range 170-270 ng/100 ml). These concentrations were far in excess of those which would have been expected on the basis of the T3 contamination of the administered T4, as measured by the same technique employed in the analysis of serum. When oral medication was enriched with (125)I-labeled T4 for 8 or more days, labeled T3 and tetraiodothyroacetic acid (Tetrac or TA(4)) were found in the serum to the extent of approximately 2-5% of total radioactivity, as assessed by unidimensional paper chromatography. The same results were obtained with a specially purified lot of radioactive T4 containing less than 0.1% T3 as a contaminant. The identities of the (125)I-labeled T3 and TA(4) were verified by two-dimensional chromatography as well as by specific patterns of binding in serum. The labeled T3 isolated was bound by albumin and by T4-binding globulin (TBG), but not by T4-binding prealbumin (TBPA): in contrast the labeled TA(4) was bound by albumin and TBPA, but not by TBG. To exclude the possibility that the conversion of T4 to T3 was a peculiarity of the oral route of administration, the sera of two additional patients were obtained 48 hr after 7-day courses of daily intravenous injections of a mixture of stable and (125)I-labeled T4. Both stable and labeled T3 were likewise found in these sera. In contrast to earlier experiments in humans in which (131)I-labeled T3 was not definitively demonstrated in serum after a single intravenous injection of (131)I-labeled T4, the present findings are taken to provide conclusive evidence of the extrathyroidal conversion of T4 to T3 in man. These results raise once again the question of the extent to which the metabolic effect of T4 is mediated through the peripheral generation of T3.