Administration of a moderate dose (2 g/kg) of ethanol to the pregnant rat during the last days of gestation (17-20) has been observed to increase the offspring's ethanol consumption on postnatal day 14. This effect was not observed with a 1-g/kg dose (Experiment 1a). When pups were tested during adolescence (day 28) an augmented ethanol intake was observed in female rats exposed to the low dose of ethanol and in male rats exposed to the moderate one (Experiment 1b). The effect of increased ethanol intake in preweanling rats was reduced after naloxone was administered, together with ethanol, to the pregnant dam (Experiment 2). Postnatal reexposure to ethanol, together with a naloxone injection, also decreased ethanol intake in pups exposed prenatally to the drug (Experiment 3). All these results seem to support the hypothesis of a conditioned preference learned in utero as a consequence of the association between the orosensory characteristics of ethanol and its reinforcing properties, apparently mediated by the opioid system.