Corticotropin releasing factor (CRF), a neuropeptide secreted by hypothalamic and extrahypothalamic neurons, is thought to mediate stress-related behaviors. The tension reduction hypothesis suggests that ethanol drinking reduces stress; that drinking is reinforced by this reduced stress; and that the probability of drinking therefore subsequently increases. CRF also decrease food intake, and might decrease ethanol drinking similarly. We addressed these hypotheses directly by assessing the effects of intracerebroventricular (i.c.v.) CRF upon ethanol drinking (1 h/day). Rats were provided drinking tubes containing ethanol solutions that were gradually incremented in concentration (from 2% to 8% w/v, over 38 days). Ethanol intakes remained stable, ranging from 0.4 to 0.5 g/kg per hour on average, and a two-bottle choice test revealed that ethanol was preferred reliably to water. Third-i.c.v. cannulae were surgically implanted and CRF or vehicle was acutely injected immediately prior to the sessions. CRF dose-dependently reduced ethanol intake by 31% (0.5 microg) and 64% (5.0 microg), and reduced 24-h food by 9% and 21%, respectively, but did not alter body weights. I.c.v. CRF reduced ethanol drinking despite any acute stress-like effects that may have been present. Hence, these data are inconsistent with the tension reduction hypothesis. On the other hand, our results support the concept that food intake and ethanol drinking may be mediated by similar mechanisms.