A method for the rapid determination of thyroxine (T4) in serum was described in 1964 (J Clin Endocr 23: 187) and modified in 1965 (J Lab Clin Med 66: 161). Since the intact thyroxine molecule is measured, the determinations are entirely unaffected by iodine of mercury. Clinical information was obtained in 1439 cases in whom the T4 had been determined. In 400 of these the PBI had also been done. A mean value of 6.36 [mu]g/100 ml was obtained in euthryoid men and 6.60 [mu]g/100 ml in euthyroid women. Adopting a mormal range of 4.0-11.0 [mu]g/100 ml, an overall compatibility with the final diagnosis of 97% was obtained in series of consecutive determinations when pregnant subjects and those receiving drugs known to alter the serum T4 [estrogen, Dilantin, T3 (tri-iodothyronine) T4] were excluded. False negatives were negligible (0.24%). Of the false positives, half of the low values (1% of the total) were accounted for by hypoproteinemia and occurred more commonly in men. All of the false positive high values occurred in women. When T4 and PBI (protein-bound iodine) values were compared, a good correlation (r = 0.823) was obtained with a higher diagnostic accuracy for the T4 determination. All euthyroid individuals with PBI''s elevated due to iodine had T4 values in the normal range. Among subjects receiving drugs, it was shown that the T4 fell with Dilantin and T3 therapy, and rose with D-thyroxine or estrogen. The T4 level correlated well with the clinical status in hypothyroid subjects receiving T4 or hyperthyroid subjects receiving various forms of therapy.