The discriminative stimulus effects of centrally administered morphine were characterized in rats trained to discriminate 3.0 mg/kg SC morphine from saline in a two-choice discrete-trial avoidance paradigm. The intracerebroventricular (ICV) administration of 0.3-10 micrograms morphine engendered morphine-appropriate responding, morphine administered ICV being nearly 1000 times as potent as morphine administered SC. Cannula implantation itself did not affect the sensitivity of the rats to the discriminative effects of morphine. The onset of the discriminative stimulus effects of ICV morphine was not immediate; stimulus generalization comparable to that produced by 3.0 mg/kg morphine occurred 30-60 min after the injection of 1.0 or 10 micrograms ICV morphine and persisted for 90 and 150 min, respectively. Naltrexone blocked the discriminative stimulus effects of 10 micrograms ICV morphine in a dose-related manner. Complete antagonism of the stimulus effects of this dose of morphine was obtained with 0.01-0.03 mg/kg SC naltrexone. When administered centrally, the relatively lipid insoluble naltrexone methobromide completely antagonized the discriminative effects of 3.0 mg/kg morphine at a median effective dose of 0.3 micrograms. In contrast, when injected systemically at a dose of 1.0 mg/kg (approximately 500 micrograms), naltrexone methobromide failed to block the discriminative stimulus effects of either 10 micrograms ICV morphine or the SC training dose. Thus, periventricular brain sites appear to be involved in mediating the discriminative stimulus effects of morphine in the rat.