BACKGROUND The conditions under which naltrexone reduces ethanol consumption and its effect on behavior controlled by ethanol conditioned stimuli remain unclear. OBJECTIVE The objectives were to determine the effects of naltrexone on ethanol consumption by C57BL/6 (B6) mice when injected subcutaneously (expt 1) or delivered by osmotic minipump (expt 2), and on ethanol conditioned cues (expt 3). METHODS Naltrexone effects on ethanol consumption and preference were measured in a continuous access two-bottle choice paradigm in groups of mice implanted with osmotic minipumps delivering 0-3.0 mg/kg per day or injected subcutaneously with 0-6.0 mg/kg doses. Naltrexone's (0-3.0 mg/kg) effect on ethanol-conditioned cues was indexed by its effect on the expression of ethanol place conditioning (expt 3). RESULTS Naltrexone produced a transient reduction in ethanol consumption and a consistent reduction in preference when injected; however, it had no effect on ethanol consumption or preference when delivered continuously by osmotic minipump. Naltrexone attenuated the expression of ethanol place conditioning in a U-shaped dose-response function. CONCLUSIONS The transient reduction in ethanol consumption produced by injected naltrexone and the absence of an effect when continuously delivered confirms a report that maintaining naltrexone at steady state levels may antagonize its attenuation of ethanol consumption. The reduced expression of ethanol place conditioning in naltrexone-injected mice suggests that the drug can attenuate the reinforcing effects of ethanol conditioned stimuli as was recently reported for lever responding maintained by ethanol conditioned stimuli in rats. These effects were observed at naltrexone doses with no readily apparent adverse side-effects, supporting its usefulness for treating alcoholism.