Heat stress is accompanied by a decrease in basal metabolic rate and plasma thyroid hormones. Unlike 3,5,3'-triiodothyronine (T3) and thyroxine (T4), 3,3',5'-triiodothyronine (rT3) displays hypometabolic properties and antagonizes the hypermetabolic effect of T3. This study analyses the role of rT3 in heat (38-39 degrees C) stressed immature chickens. Two experiments which differed in frequency of rT3 injections (one or two times a day), duration of heat stress (72 or 48 h) and blood sampling were performed. The dose was 14 micrograms rT3/100 gb.wt./injection (s.c.). It has been shown that rT3 treatment aggravates heat stress symptoms (enhances circulating corticosterone, catecholamines and free fatty acids) and increases mortality. The critical survival time of the rT3 treated and heated birds was at first 24 h of stress. No more chickens died during the next days of the experiment despite the continuation of rT3 injection, suggesting that rT3 might disturb the adaptation to heat. Reverse T3 in heat stressed chickens led to the highest reduction in food consumption (69.9%) and body weight gain (14.0% compared to initial weight). The opposite effect in water consumption (216.9%) was observed. In a neutral environment, rT3 significantly suppressed body temperature 6 h after injection (40.4; control; 41.1 degrees C), confirming its hypometabolic properties. However, at the same time rT3 significantly enhanced body temperature in heat stress (43.03 versus heated control 42.56 degrees C). In addition, in rT3 treated birds decreased plasma triglycerides (TG; 24.3%) and increased plasma free fatty acids (FFA; neutral temperature; 26.4% heat stress: 57%) were demonstrated. A correlation between corticosterone and FFA (r = 0.52) shows that some of the FFA may originate from lipolysis since hormones of the pituitary-adrenocortical axis accelerate lipolysis. The remaining part of the increased FFA appears to be due to suppressed utilization of FFA as a consequence of hypometabolic properties of rT3. Low and negative relation between TG and FFA (r = -0.26; P < 0.05) may support such an assumption. The two times higher peak of corticosterone in the rT3 and the overheated group, as compared to the heated control, occurred at 6 h of heat stress and indicates that rT3 increases the unfavourable effect of high temperature. This was also confirmed by elevated plasma adrenaline and noradrenaline in rT3-injected and heated chickens (55.5 and 120%, respectively). However, a single and two times higher peak of adrenaline at 24 h of heat stress was observed in saline treated birds, but not in rT3 supplemented animals, suggesting that this difference might explain one of the factors responsible for high mortality. In conclusion, the results obtained demonstrate that physiological doses of rT3, a hypometabolic hormone, enhance the unfavourable effect of heat stress in chickens.