The peripheral injection of alcohol stimulates the activity of the hypothalamic-pituitary-adrenal (HPA) axis, but the ready penetration of this drug in most bodily compartments has made it difficult to identify its specific sites of action. Here we determined whether alcohol acts directly on the corticotrophs. We first determined whether alcohol stimulated neurons in the paraventricular nucleus (PVN) of the hypothalamus, that synthesize corticotropin-releasing factor (CRF) and vasopressin (VP). To test this hypothesis, we injected alcohol intracerebroventricularly (icv; 5 µl of 200 proof; 86 µmole) and compared these results with those obtained following its intraperitoneal (ip) administration (3.0 g/kg). While not causing neuronal damage and not leading to detectable levels of the drug in the general circulation, icv alcohol significantly upregulated PVN CRF heteronuclear RNA levels and increased plasma ACTH levels, a change comparable to the one observed in the ip model. To determine whether alcohol stimulated the corticotrophs independently of CRF and/or VP, we injected the drug ip or icv and measured changes in anterior pituitary POMC transcripts and ACTH release in the presence or absence of endogenous CRF and/or VP. Icv and ip alcohol significantly increased POMC primary transcript levels, measured by RNase protection assay, over a time-course that corresponded to ACTH release. Both the POMC and the ACTH responses were completely abolished by removal of CRF and VP. Collectively, these results indicate that alcohol-induced activation of the corticotrophs does not represent a direct influence of the drug on the pituitary, but requires CRF and VP.