T4 and reverse T3 (rT3) can inhibit 5'-deiodinase type II activity in rat brain cortex, pituitary, and brown adipose tissue, raising the possibility that T4 may act in vivo after conversion to rT3. The aim of this study was to measure in hypothyroid (Tx) rats the content of brain cortex rT3 during a constant 7-day infusion of either [125I]T4 alone, corresponding to 12 pmol T4/day X 100 g body weight (BW), or together with 400 pmol T4/day. [125I]T4, rT3, and T3 were extracted from brain cortex, pituitary, kidney, and liver with a combination of adsorption chromatography on Sephadex G-25, HPLC, and immunoprecipitation. [131I]T4, T3, or rT3 were used as internal standards. [125I]rT3 could be detected in brain cortex, liver, and kidney in Tx rats infused with [125I]T4 (12 pmol T4/day X 100 g BW) and in those infused with 400 pmol T4/day X 100 g BW. The highest rT3 concentrations were found in brain cortex, where it represented 6% to 10.5% of the local T4 concentration. During an infusion of 400 pmol T4/day X 100 g BW, brain cortex T3 concentration was 6 times higher in the brain cortex than in serum, and even exceeded that of T4. In Tx rats receiving [125I]T4 alone the brain cortex to serum T3 ratio was 3:1, but the total serum T3 concentration, measured by RIA, was much higher than that due to conversion [0.50 +/- (SE) 0.1 pmol/ml vs. 0.018 +/- 0.002 pmol T3/ml], indicating thyroidal secretion. The effect of the blood-brain barrier on rT3 was measured by infusing [125I]rT3 over 4 days. After killing, rT3 was isolated as above. Approximately 3% of serum rT3 was retrieved from the brain cortex, whereas during the T4 infusion 40-50% of serum rT3 was found demonstrating that brain cortex rT3 is locally produced.