Amiodarone, an iodinated benzofuran derivative, has electrophysiologic effects on cardiac muscle akin to those of hypothyroidism. It is possible that the drug exerts its salutary effect, at least in part, by selectively inhibiting the action of triiodothyronine (T3) on the myocardium. The drug produces complex changes in thyroid hormones, with significant elevations in thyroxine (T4) and reverse T3 (rT3), with minor decreases in T3, and with minor and transient increases in thyroid-stimulating hormone, but without effect on thyroid-binding globulin. These changes may interfere with the biochemical evaluation of thyroid function. Rarely, hypothyroidism or hyperthyroidism may develop during the course of amiodarone therapy, a complication caused by the iodine contained in the drug rather than by the direct pharmacologic actions of the compound. The incidence of altered thyroid function induced is likely to vary with populations susceptible to iodine-induced goiter. Under the action of amiodarone, serum rT3 levels increase as a function of dose and duration of therapy and therefore provide a basis for judging the magnitude of in vivo drug cumulation. It was found that therapeutic efficacy was usually predictable on the basis of the attainment of a defined range of serum values, established by a correlation of rT3 levels with therapeutic responses both during loading and maintenance phases as well as after withdrawal of treatment of steady-state drug effects. Serious adverse effects occurred nearly always in association with four- to fivefold increases of rT3 above baseline values, and disappeared when such levels fell as a result of dosage reduction or after temporary drug discontinuation. The data suggest that the determination of serum rT3 levels during amiodarone therapy provides a simple and reliable technique for monitoring the drug's antiarrhythmic efficacy and toxicity, thereby enhancing its clinical utility. The use of rT3 levels may permit the development of a safe but optimal therapeutic regimen for the control of a wide spectrum of refractory atrial and ventricular tachyarrhythmias. The use of this technique, however, presupposes the allowance that must be made for variations in the methods for the serum assay of rT3 and of the systemic conditions in which the rT3 levels fluctuate relative to severity of the illness.