We treated 204 patients with endemic nontoxic goiter with T4, T3 and KI, singly or in combination. Definitely nodular goiters were excluded, since the possibility of autonomy would be increased. Goiter size was evaluated before and 6 months after treatment clinically in a blind way, i.e. the observer (always the same) did not know either the pretreatment goiter size or the treatment the patient had received. At the same time various laboratory parameters were recorded. All the active treatments (but not placebo) resulted in a highly significant decrease in the gland size. The effectiveness decreased in the following order: 1) T3 50 micrograms/d (most effective), 2) (T4 50 micrograms/d + T3 12.5 micrograms) x 2, 3) T4 150 micrograms + iodide 150 micrograms/d, 4) T4 75 micrograms + T3 18.75 micrograms/d, 5) T4 200 micrograms/d, 6) T3 37.5 micrograms/d, 7) Iodide 300 micrograms/d, 8) T4 150 micrograms/d, 9) Iodide 150 micrograms/d (least effective) and 10) Placebo (not effective). The results show that T4 200 micrograms and T3 50 micrograms are roughly equipotent, and slightly more effective than 300 micrograms of Iodide. Taking into consideration the side effects (increase in pulse rate, shortening of the Achilles tendon reflex) did not change the order of effectiveness in an important way. The clinical outcome correlated in general with the suppression of the 131I uptake (r = 0.220, p = 0.03) and the TRH test (r = 0.248, p = 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)