Studies were designed to examine the previously proposed hypothesis that some of the pharmacological actions of ethanol are mediated by neurotensinergic processes. Neurotensin-immunoreactivity (NT-ir) was extracted from various brain regions and shown by high performance liquid chromatography to possess the same retention time as authentic bovine NT1-13. The highest levels of NT-ir were observed in the hypothalamus with intermediate levels in the midbrain and striatum and lowest levels in the frontal cortex. Levels of NT-ir were higher in hypothalamus and midbrain from long-sleep (LS) than from short-sleep (SS) mice. Ethanol, in vivo, produced a dose-dependent decrease in NT-ir in several brain regions; low doses, 1.5 to 3.0 g/kg, but not high doses, 4.1 g/kg, of ethanol significantly decreased NT-ir in hypothalamus, midbrain, and striatum of LS and SS mice. Levels of NT-ir in the frontal cortex were not altered by ethanol administration. Ethanol-induced decreases in NT-ir were of rapid onset with a maximum decrease in 5 min after intraperitoneal (i.p.) injection, and they were of long duration with levels remaining depressed for 4 hr. These findings show that subhypnotic, intoxicating doses of ethanol enhance NT release, in vivo, and support the hypothesis that some of ethanol's actions are mediated by neurotensinergic systems.