The importance of local formation of T3 in the feedback effect of the thyroid gland on hypothalamic TRH-producing cells has been established. Primary failure of the thyroid gland results in a fall in circulating T4 and T3 levels, leading to an elevation in the production and release of TRH in the hypothalamic paraventricular nucleus. In contrast, during short term fasting, declining plasma levels of thyroid hormones coincide with suppressed TRH production and release. In the brain, the prevalent enzyme that converts T4 to T3 is type II iodothyronine deiodinase (DII). The present study was undertaken to determine whether a differential hypothalamic expression of type II deiodinase may exist in fasted rats and in animals that are hypothyroid due to the failure of the thyroid gland. Using in situ hybridization, we assessed type II deiodinase messenger RNA (mRNA) levels in the hypothalamus of rats that were control euthyroid, hyperthyroid (T4), hypothyroid induced by propylthiouracil (PTU), and fasted. A group of fasted rats also received exogenous T4. DII mRNA was detected around the third ventricle, including the ependymal layer and adjacent periventricular regions as well as in the arcuate nucleus and the external layer of the median eminence. Quantitative in situ hybridization analysis demonstrated that PTU treatment and short term fasting resulted in significant elevations in DII messenger levels compared with those in euthyroid controls. Three weeks of PTU administration induced a consistent decline in circulating T3 and undetectable T4 levels, whereas 3 days of fasting resulted in only a 50% fall in the concentration of serum thyroid hormones. Interestingly, however, the expression of the DII mRNA was more than 2-fold higher in fasted animals compared with the values in PTU-treated rats. Furthermore, although T4 administration repressed DII mRNA expression in euthyroid animals, the same treatment had no effect on the fasting-induced elevations of DII message. To assess whether DII enzymatic activity is also affected during food deprivation, hypothalami were dissected out, and DII activity was measured in control euthyroid, fasted, and fasted plus T4-treated rats. To determine whether comparable changes in plasma thyroid hormone levels induced by fasting and PTU treatment could have affected DII enzymatic activity in a similar manner, animals were injected ip with PTU for 5 days to decrease plasma thyroid hormones to levels similar to those caused by fasting. DII enzymatic assay showed a significant increase in DII activity in fasted and fasted plus T4-treated animals compared with those in euthyroid controls and PTU-treated rats. No significant changes were found in PTU-treated rats compared with euthyroid animals. These data indicate that during short term fasting, a signal of nonthyroid origin underlies the robust elevation of DII production and activity in the hypothalamus. Thus, we propose that during the initial phase of food deprivation, an increased negative thyroid feedback exists on the hypothalamus due to locally formed T3. This local hyperthyroidism may, in turn, induce the suppression of TRH under these conditions.