From the foregoing discussion, it is clear that no single test provides sufficient information to justify its use alone as a single screening test. In vitro tests have now replaced in vivo procedures in the vast majority of patients. Because of the frequency of abnormalities in TBP concentration, the estimation of total T4 should be accompanied by a T3 resin uptake to provide the free thyroxine index or alternatively, a normalized T4 test (Quantisorb or ETR) is preferable. In patients with suspected hyperthyroidism, the initial laboratory evaluation should be an estimate of free T4 and a total serum T3 determination. Whereas the majority of hyperthyroid patients exhibit elevated free T4 levels, a smaller but variable percentage will exhibit only an elevated T3 level. The diagnosis mients where equivocal tests do not provide a diagnosis. In patients with suspected hypothyroidism, estimations of T4 and T3 provide evidence of diminished thyroidal secretion. The diagnosis should be confirmed by demonstration of an elevated TSH level. Normal or low TSH levels point to a diagnosis of pituitary hypothyroidism which can be confirmed by TRH stimulation. The finding of low normal or subnormal T4, normal T3 and elevated TSH levels suggest "compensated hypothyroidism". Estimation of thyroid autoantibodies may confirm the diagnosis of autoimmune thyroiditis. It is emphasized that the approach to testing thyroid function should be an adequate clinical assessment so that selection of the appropriate test(s) currently available leads to a diagnosis of great certainty in most cases.