We report here a study of the effects of alprazolam on in vivo pituitary-adrenal function in jacketed nonrestrained nonhuman primates and on in vitro CRH release from rat hypothalami and ACTH release from rat dispersed anterior pituicytes. We undertook this study because alprazolam is the only benzodiazepine effective in treating both major depressive and anxiety disorders, and recent data suggest that the hypercortisolism of major depression reflects hypersecretion of CRH. Moreover, the intracerebroventricular administration of CRH can reproduce many of the components of the symptom complex of major depression, including not only hypercortisolism, but also hypothalamic hypogonadism, decreased libido, anorexia, and intense anxiety. As a comparison, we also assessed the effects of diazepam on in vitro CRH release, because in contrast to alprazolam, diazepam is effective in anxiety states but not in depression. Alprazolam (0.01-0.3 mg/kg, iv) produced a dose-dependent inhibition of both plasma ACTH and cortisol secretion in non-restrained adult male rhesus monkeys. Our in vitro studies showed that alprazolam significantly inhibited serotonin (5HT)-induced CRH release in a dose-dependent fashion (10(-10)-10(-5) M). Diazepam also inhibited 5HT-induced CRH release, but was 40 times less potent than alprazolam. Alprazolam was ineffective in blocking basal or CRH-induced ACTH release from rat dispersed anterior pituicytes, suggesting that its in vivo effects are through inhibition of CRH secretion. As expected, the inactive benzodiazepine ligand Ro 15-1788 inhibited the effects of alprazolam on 5HT-induced CRH release, but this occurred only at doses below 10(-7) M. Interestingly, when incubated alone in higher doses with our rat hypothalamic organ culture, Ro 15-1788, like alprazolam, produced a dose-dependent inhibition of 5HT-induced CRH release (10(-7)-10(-5) M), suggesting an agonistic action of Ro 15-1788 at the benzodiazepine receptor at higher concentrations. We conclude that alprazolam is capable of suppressing the primate pituitary-adrenal axis, and that this suppression most likely reflects suppression of the CRH neuron rather than of the pituitary corticotroph cell. We speculate that the enhanced capacity of alprazolam to suppress the CRH neuron relative to other benzodiazepines may contribute to its unique efficacy among this class of drugs in the treatment of major depression. The capacity of Ro 15-1788 to reverse alprazolam-induced suppression of the CRH neuron indicates that the effects of alprazolam are mediated at least in part via its interaction with the benzodiazepine component of the gamma-aminobutyric acidA macromolecular complex.