Serum levels of T4, T3, rT3, RT3U and TSH were estimated in 12 obese women in the course of a 14-day fasting. Seven of these patients were treated with T3 in a daily dose of 60--80 micrograms. Fasting led to a small increase of serum T4, while fasting combined with T3 administration was accompanied by a small decrease of serum T4. Serum T3 decreased in the course of the first 2 days of fasting to 40--50% of initial values and remained at this low level up to the 15th day of fasting. In the T3 treated group a prompt increase of serum T3 was recorded (+80%), followed by a steady decrease, reaching the control values on day 15, in spite of a continuous T3 administration. Serum rT3 in the untreated fasting group steadily increased up to the 12th day (+43%) and then dropped below the control value (-28%). In the treated group after a non-significant increase (+22%), a decrease of serum rT3 was also observed (-42%). Fasting was accompanied by a tendency to increased serum RT3U values, in the T3 treated group no change occurred. TSH in the untreated fasting women remained practically unchanged while T3 administration was accompanied by a tendency to a transient decrease. The heart rate showed a non-significant tendency to increase in the T3 treated group and the last week of fasting was also accompanied in several patients by a larger weight loss. These results suggest that in addition to an enhanced synthesis of rT3 to the detriment of a decreased production of T3 from T4, an increased catabolism of administered T3 and of endogenous rT3 is involved in a strict calorie restriction. An abortive T3-TRH-TSH feed-back may also operate in these conditions.