Chromatography of serum on columns of Sephadex G-25 superfine after the iv administration of 125I-labeled T4 consistently yielded labeled iodide, iodoprotein, T3, and a labeled peak that eluted from the column before the T3, termed "pre-T3." Much larger quantities of pre-T3 were generated after the iv administration of 125I-labeled T3 and rT3. The ratio of [125I]pre T3:[125I]T3 and [125I]pre T3:[125I]rT3 plateaued at approximately 10 h and 2 h, respectively, after the iv administration of the labeled hormone, and averaged approximately 15% in euthyroid subjects. As pre-T3 behaves like its precursors in the TCA precipitation-ethanol extraction or anion exchange chromatographic procedures used to separate labeled iodide and iodoprotein from administered labeled T3 and rT3, concentrations of labeled hormone measured by these techniques will be in error, because pre-T3 will be included. Thus, the MCR of labeled T3 and rT3 measured by Sephadex chromatography will always be higher than values obtained by the other two separative techniques and the magnitude of change will be similar to the ratio of pre-T3 to precursor T3 or rT3. The Sephadex chromatographic technique is laborious, but appears to be the method of choice where greatest accuracy of measurement is required. As the generation of pre-T3 from labeled T4 is sufficiently slow, the present chromatographic technique is not necessary in studying peripheral T4 turnover.