Recent studies have indicated that there is rapid in vivo conversion of L-[¹²⁵I]T₄ to L-T₃ in rat cerebral cortex and cerebellum. These early studies showed significant quantities of [¹²⁵I]T₃ specifically bound to nuclear thyroid hormone receptors. The amount of [¹²⁵I]T₃ relative to the simultaneously injected [¹²⁵I]T₃ in cell nuclei was greater in the cerebral cortex and cerebellum than in other tissues examined, suggesting a major contribution of locally generated T₃ to total nuclear T₃ in these organs. The present studies have been performed to assess the relative contribution of both T₃ sources, the plasma and local 5′-deiodination to the total mass of T₃ specifically bound to the nuclear receptors of cerebral cortex and cerebellum in euthyroid rats in steady state conditions. To determine the maximal binding capacity of the nuclear receptors in these two tissues, rats were injected with increasing quantities of unlabeled T₃ along with labeled T₃. The maximal binding capacity of the nuclear receptors in cerebral cortex was 266 pg T₃/mg DNA (95% confidence limits, 242–290) and that for cerebellum was 95 pg T₃/mg DNA (89–101). Euthyroid rats were then given [¹²⁵I]T₄ and [¹³¹I]T₃ and, after suitable corrections for the contribution of plasma [¹²⁵I]T₃ to nuclear [¹²⁵I]T₃, they were given the locally generated nuclear [¹²⁵I]T₃. Local (Lc) T₃(T₄), was computed. At 15 h or longer after [¹²⁵I]T₄ injection, Lc T₃(T₄) accounted for more than 80% of the total nuclear [¹²⁵I]T₃ in cerebral cortex and approximately 67% of that in the cerebellar nuclei. The nuclear Lc [¹²⁵I]T₃(T₄) to serum [¹²⁵I]T₄ ratio was constant between 15 and 24 h after the injection. The gravimetric amount of nuclear Lc T₃(T₄) was thus estimated by multiplying this ratio by serum T₄ RIA. In addition, the nuclear to serum ratio for T₃, at the peak of nuclear [¹³¹I]T₃, and the serum T₃ RIA were used to estimate the contribution of plasma T₃ to nuclear T₃. In the cerebral cortex, nuclear Lc T³(T₄) was 195 pg/mg DNA occupying 73% of the maximal binding capacity of these receptors. Plasma T³ contributed 56 pg T³/mg DNA, 21% of the binding capacity, resulting in over 90% occupancy of the nuclear thyroid hormone receptors. For cerebellum, Lc T³(T⁴) was 30 pg/mg DNA (32% of the maximal binding capacity), whereas plasma T³ contributed about 23 pg/mg DNA or 24% of the maximal binding capacity. The occupancy of cerebellar nuclear thyroid hormone receptors was 56%. These results indicate that T³ generated intracellularly from T⁴ contributes more than half of the T³ specifically bound to nuclear receptors of cerebral cortex and cerebellum. Furthermore, it appears that these receptors in cerebral cortex are nearly saturated at endogenous serum T³ and T⁴ concentrations. The results suggest that T⁴ plays an important role as a source of intracellular T³ in the central nervous system.