Subcutaneous injection of 400 µg/100 g body weight of corticosterone (B) 2 h previously in male rats prevented the stress response, as assessed by the ability of adrenal glands removed from these animals to produce endogenous B. Two injections of a combination of 17α-hydroxyprogesterone and 11-epicortisol, the first given 30 min before and the second with the B, were able to block this inhibitory effect on the stress response. Neither of the steroids alone was effective in this regard. The combination was also effective against the early delayed feedback effects of 400 µg/100 g body weight cortisol, prednisolone or beclomethasone dipropionate in the same system. The minimum effective dose for reversal of feedback by B or beclomethasone dipropionate (2 mg/100 g body weight of each antagonist) was lower than that required for the same effect against prednisolone or cortisol (5 mg/100 g body weight). Previous injection of B also abolished the ability of anterior pituitary gland fragments to respond to corticotropin-releasing factors (CRFs) added in vitro, an effect which was not abolished by the injection of the combination of putative antagonistic steroids. From experiments designed to measure the ability of 17α-hydroxyprogesterone and 11-epicortisol to compete with <sup>3</sup>H-corticosterone in binding to macromolecular components in hypothalamic, hippocampal and pituitary cytosolic preparations, it was deduced that the competition seen in the hypothalamic and hippocampal, rather than the pituitary, preparations was in better accord with the effect seen on the stress response. These results, taken together with our studies of the effects of the two steroids on CRF release from the hypothalamus in vitro, suggest that the site of action of 17α-OH-progesterone and 11-epicortisol is at the hypothalamus (and/or higher centres) and not at the pituitary gland.