Six experiments were carried out to determine whether dihydrotestosterone (5-alpha-androstan-17-beta-ol-3-one; DHT) acts to inhibit oestradiol (OE2)-induced lordosis behaviour after its metabolic conversion to 5-alpha-androstane-3-alpha,17-beta-diol (3-alpha-androstanediol, 3-alpha-Adiol). In experiments 1 and 2, ovariectomized rats were treated with several doses of DHT or 3-alpha-Adiol, injected with OE2 and progesterone, and tested for lordosis responsiveness. Significant inhibition of lordosis occurred after a dose of 3-alpha-Adiol which was approximately threefold less than the effective DHT dose. In experiments 3 and 4, plasma concentrations of DHT and 3-alpha-Adiol were measured after the injection of these steroids to ovariectomized rats at doses shown to be both sufficient or insufficient to inhibit lordosis. Behaviourally effective dosages of DHT and 3-alpha-Adiol produced circulating concentrations of 3-alpha-Adiol greater than those produced by behaviourally ineffective doses of DHT or 3-alpha-Adiol. At 30 min after injection of DHT (experiment 3), 78.8% of plasma androgens were in the form of 3-alpha-Adiol, while after injection of 3-alpha-Adiol, only 7.4% were DHT. When plasma DHT and 3-alpha-Adiol were measured at 3, 6, 9 and 12 h after steroid injection (experiment 4), plasma levels of 3-alpha-Adiol produced by the behaviourally subthreshold dose of DHT were significantly lower than levels produced by behaviourally sufficient dosages of DHT or 3-alpha-Adiol. In experiments 5 and 6, concentrations of DHT and 3-alpha-Adiol were measured in five brain regions 1 and 6 h after injection of behaviourally sufficient doses of these steroids to ovariectomized females. At 1 h after injection, similar levels of DHT and 3-alpha-Adiol were measured in DHT- and 3-alpha-Adiol-injected females, and DHT concentrations in the preoptic area were significantly higher in both groups than in any other brain area. At 6 h, animals injected with DHT had significantly higher levels of DHT in all brain areas combined than did 3-alpha-Adiol-or vehicle-injected animals. Brain concentrations of 3-alpha-Adiol were not different between groups injected with DHT, 3-alpha-Adiol or vehicle at this time. In brain, 34.6% of DHT had been converted to 3-alpha-Adiol after 1 h and 53.0% of 3-alpha-Adiol had been converted to DHT. These results suggest that the inhibitory action of DHT on lordosis may be a consequence of its conversion to 3-alpha-Adiol, and that this conversion may account for the higher behavioural potency of the latter steroid.