The clinical value of four laboratory methods of assessing free thyroxine status was compared in 82 consecutive patients newly referred to a thyroid clinic with suspected thyroid dysfunction. The methods of determining free thyroxine used were: (1) free thyroxine index using a thyroid hormone uptake test (FTI (THUT)); (2) a free thyroxine index using thyroxine binding globulin (FTI (TBG)); (3) a kinetic radioimmunoassay (Immophase); and (4) an equilibrium dialysis method. The definitive thyroid status was evaluated by a combination of clinical assessment (including Wayne index), routine tests of thyroid function (total T4, T3, TSH, and TRH tests where appropriate), and by therapeutic trial in one case. The diagnostic efficiency of the tests was markedly dependent upon the method of determining the reference range for euthyroid patients. Best efficiency for each test was achieved using an amended range after excluding outliners. Test efficiency was then 97.6% for FTI (THUT) and the kinetic RIA and 96.8% for FTI (TBG). Misclassification by one or more of these tests occurred in only four patients (mild hypothyroid, euthyroid on phenytoin, euthyroid on oral contraceptive and valium, T3 hyperthyroid). In contrast, free T4 by equilibrium dialysis was much less efficient (86.6%) and was technically the most complex. Overall the kinetic T4 RIA provided similar diagnostic information to the indirect indices. However, further studies of cost benefit in settings other than a thyroid clinic are required to assess whether this method might replace total T4 and/or FTI as a first-line test of thyroid function.