Eight open clinical trials conducted by 7 different study groups and including 78 patients have all demonstrated that augmentation with supraphysiological doses of thyroxine (T4) has antidepressant and prophylactic effects in roughly 50% of patients completely resistant to all other antidepressant and prophylactic therapies. Beneficial effects have been observed in unipolar and bipolar (rapid-cycling and non-rapid-cycling) patients, but only when an antidepressant or prophylactic drug was administered concomitantly. Double-blind studies are now needed in order to confirm these results. It has also consistently been shown that high serum concentrations of T4 predict favourable response to antidepressant treatment and that the serum levels of T4 decrease in responders to these treatments, but not in non-responders. As thyroid hormone function in the CNS depends almost entirely on the uptake of T4 and its intracellular deiodination to the active compound T3, the hypothesis was investigated that the falls in serum levels of T4 seen during antidepressant treatment are due to enhanced conversion of T4 to T3 in the CNS. However, the results of several animal studies revealed that, while a number of different antidepressants do in fact each have specific effects on thyroid hormone metabolism in the CNS, no consistent enhancement of T3 concentrations has been demonstrated in homogenates of any relevant brain region. Recent studies measuring T3 in subcellular fractions have reported a selective increase in T3 levels in the mitochondria of the amygdala following various antidepressant treatments. The relevance of this finding must be clarified in further studies. However, in humans serum levels of T4 also decline after non-antidepressant treatments (for example, neuroleptics, anticonvulsants or benzodiazepines), and T3 concentrations in the rat brain are elevated by many other kinds of non- antidepressant treatment (e.g. stress). The function of T3 appears to be a rather general enhancement of all kinds of neuronal activity. Thus, it would seem unlikely that effects on thyroid hormone function are the decisive and specific step involved in the mechanism of action of antidepressant treatments. Rather, the function of T3 is altered as a secondary response to other primary effects of antidepressant treatments and also other psychopharmacological therapies.