The activities of the 5'I-deiodinase (5'D-I), 5'II deiodinase (5'D-II) and 5III-deiodinase (5D-III) isoenzymes and tissue concentrations of thyroxine (T4) and triiodothyronine (T3) were measured in up to 10 regions of the rat brain after acute and subchronic nonpharmacological (sleep deprivation, 12 h fasting, 14 days' calorie-reduced diet) and pharmacological (ethanol, haloperidol, clozapine, lithium, carbamazepine, desipramine, fluoxetine, tranylcypromine, and mianserin) treatments. All of these treatments induced significant and sometimes dramatic changes in 5'D-II activities and tissue concentrations of thyroid hormones and, to a lesser extent, in 5D-III activity. The activity of 5'D-I remained unaffected. The results revealed a surprising specificity for each type of treatment in terms of the isoenzyme and hormone affected, the direction of the change, the brain region affected and the time of day. The changes in thyroid hormone concentrations frequently failed to correspond in any way to those in deiodinase activities and unexpected effects such as inhibition of both 5'D-II and 5D-III were seen, indicating that there may be additional pathways of iodothyronine metabolism in the CNS. In conclusion, particularly 5'D-II activity and thyroid hormone concentrations in the CNS are highly sensitive to many different kinds of influence that may induce changes in neuronal activity. However, these changes in deiodinase activities do not ensure stable tissue concentrations of T3, but were, on the contrary, in most cases accompanied by marked changes T3 levels in the tissue. The implications of these findings for the physiological role of thyroid hormones in the CNS are discussed.