In (28- to 40-day-old) male rats, the influence of castration or hormone-replacement treatment with dihydrotestosterone (5alpha-DHT) upon pituitary or hypothalamic 5alpha-reductase was assessed. The efficacy of the treatments was demonstrated by measuring serum LH and ventral prostate weight. Pituitary 5alpha-reductase enzyme activity was estimated by isolating [5alpha-(3)H]DHT by thin-layer chromatography after incubations with [(3)H]testosterone, and the identity of the [5alpha-(3)H]DHT formed was confirmed by recrystallization experiments. Pituitary and hypothalamic 5alpha-reductase mRNA content was determined by RNA blot analysis utilizing a rat 5alpha-reductase (type 1 or type 2) cDNA as probe. Pituitary 5alpha-reductase activity was significantly increased as a function of days after castration, whereas castration plus DHT treatment (at Day 12 post-castration) significantly decreased the activity to less than one-third of control levels. In controls, pituitary 5alpha-reductase mRNA content was barely detectable using the rat 5alpha-reductase (type 1) cDNA as probe. However, castration resulted in a clear increase in mRNA abundance, while in DHT-injected castrated animals pituitary mRNA content was undetectable. Hypothalamic mRNA content was clearly detectable in all treatment groups with the rat 5alpha-reductase type 1 probe. However, there were no apparent changes in hypothalamic mRNA abundance as affected by the treatments. On the other hand, pituitary or hypothalamic 5alpha-reductase mRNA was undetectable using the rat 5alphareductase (type 2) cDNA as probe. These results indicate that in the rat pituitary the content of mRNA encoding 5alpha-reductase is negatively regulated by DHT. The increase in pituitary 5alpha-reductase enzymatic activity in castrates is due, in part, to higher mRNA levels detected by the 5alpha-reductase type 1 probe. Hypothalamic 5alpha-reductase mRNA levels (type 1) encoding this protein are not regulated by DHT, but presumably are, to some extent, under transcriptional control.