In the rat, the acute administration of alcohol induces dose-related increases in plasma ACTH and corticosterone levels. This response depends on the delivery of the hypothalamic peptides corticotropin-releasing factor and vasopressin (VP) to the pituitary. On the other hand, exposure to an alcohol diet for 7 to 10 days significantly blunts the hypothalamic-pituitary-adrenal (HPA) axis in response to other homeostatic threats, such as mild electroshocks or immune signals. This decreased response is at least in part due to an attenuated ability of VP to increase ACTH secretion. We have previously shown that nitric oxide (NO) inhibits the pituitary response to VP. We therefore hypothesized that chronic alcohol treatment might increase levels of this gas within the HPA axis, and tested this possibility by determining whether blockade of NO formation might restore a normal pituitary response to VP. We observed that such was the case, and therefore propose that NO participates in the blunted activity of the HPA axis during prolonged exposure to alcohol. Finally, we determined whether alcohol would exert neuroendocrine effects that extended beyond the initial drug treatment. To explore this possibility, we injected rats with alcohol intragastrically for 3 days, then re-exposed them to the drug 3 to 12 days later. Rats pretreated with the vehicle and injected with alcohol several days later showed the expected significant rise in plasma ACTH and corticosterone levels, as well as a marked increase in immediate early genes mRNA levels in the paraventricular nucleus (PVN) of their hypothalamus. In contrast, animals pretreated with alcohol exhibited a blunted hormonal and hypothalamic response during the second drug exposure but, interestingly, retained a normal endocrine response to other signals, such as exposure to electro-footshocks or cytokine injection. We originally thought that this phenomenon of selective endocrine tolerance might be explained by decreased serotonin levels in the PVN. Whereas alcohol indeed decreased concentrations of this neurotransmitter in the PVN, exposure to electroshocks induced similar changes. However, an initial exposure to shocks did not blunt the ability of the HPA axis to be activated by a second shock session, by alcohol or by immune signals (delivered several days later). These results do not support the hypothesis that the decreased HPA axis response to a second alcohol challenge is mediated via decreased serotonin afferents to corticotropin-releasing factor neurons in the PVN.