As judged by the response of uncoupling protein and key enzymes, brown adipose tissue (BAT) is highly dependent upon the local generation of T₃ catalyzed by the tissue type II T₄ 5′-deiodinase (5′D-II). In hypothyroid rats treated with T₃ or T₄, the capacity to withstand cold seems better correlated with the normalization of BAT responses than with the liver thyroid status. 5′D-II is activated by cold via sympathetic nervous system (SNS) stimulation, and the activation generates enough T₃ to nearly saturate BAT nuclear T₃ receptors (NTR) in euthyroid rats. In hypothyroidism, 5′D-II is highly stimulated by the SNS and hypothyroxinemia. In the present studies we have taken advantage of this situation to test 1) the capacity of 5′D-II to maintain nuclear T₃ in rats with various degrees of hypothyroxinemia, and 2) the hypothesis that thyroid hormonedependent BAT-facultative thermogenesis, rather than the effect of thyroid hormone on obligatory thermogenesis (basal metabolic rate), is the basic mechanism by which thyroid hormone confers protection against acute cold exposure. We treated methimazole-blocked rats (undetectable plasma T₄ and T₃) for a week with either subreplacement doses of T₄ (0.5, 1, 2, and 4 μg/kgday) or replacement doses of T₄ or T₃ (8 or 3 μg/kgday, respectively). Sources and content of BAT nuclear T₃ were studied at 25 C and after 48 h at 4 C by labeling the plasmaborne T₃ (T₃[T₃]) with [¹³¹I]T₃ and the locally generated T₃ (T₃[T₄]) with [₁₂₅I]T₄. Neither the kinetics of nuclear-plasma exchange of T₃[T₃], the time of appearance of T₃[T₄] in BAT nuclei, nor NTR maximal binding capacity (0.71 ng T₃/mg DNA) was affected by hypothyroidism. Kinetic analyses indicated a maximal BAT NTR occupancy of 40% at euthyroid serum T₃ concentrations if T₄ is not present. Replacement with T₄ normalized both serum T₄ and T₃, while replacement with T₃ normalized serum T₃; for all other doses of T₄, serum T₄ and T₃ concentrations were predictably related to the dose. 5′D-II activity decreased with increasing doses of T₄, but for each dose of T₄, this activity was 2–4 times greater at 4 C than at 25 C. BAT NTR occupancy normalized with 2 μg T₄/kg in rats maintained at 25 C and with 4 μg T₄/kg in cold-exposed rats, although in neither condition were serum T₄ and T₃ normalized nor more than 30% of NTR occupied by plasma T₃. The thermogenic response to cold, measured as an increase in oxygen consumption (QO₂), was significantly greater in T₄-replaced than in T₃-replaced rats. Two micrograms of T₄/kg/day sufficed to normalize the response of QO₂ to cold. Prazosin, which blunts the adrenergic- induced stimulation of 5′D-II, caused a marked reduction in T₃[T₄] and blunted the T₄-dependent increase in QO₂. We conclude that the protective effect of thyroid hormone against coldinduced hypothermia is largely the result of the stimulation of facultative thermogenesis in BAT, stimulation in which the adrenergic activation of BAT 5′D-II plays an essential role. This enzyme is, furthermore, of adaptive value in hypothyroidism by allowing a normal thermogenic response to cold in the face of substantial reductions in plasma T₄. (Endocrinology128: 2149– 2159,1991)