In a one-trial step-through passive avoidance task, pretraining administration of ethanol was shown to decrease the latencies to step through at both training (day 1) and testing (day 2) for both rats and mice. A detailed analysis of these effects showed that they differed from those reported previously by others. The mechanisms underlying these effects of ethanol were also examined. The decreased day 1 latency to step through seen in rats may have been caused by an ethanol-induced hypermotility. However, ethanol did not increase the locomotor activity of mice, although it also reduced the day 1 latency to step through of this species. In addition, it was found that the ethanol-induced impairment of passive avoidance responding (i.e. the decreased day 2 latency to step through) was not state-dependent and that it was unlikely that it could be explained by a drug-induced impairment of task acquisition, long-term memory formation or memory recall. It also seemed unlikely that the impairment could be explained by an ethanol-induced decrease in shock sensitivity. Other mechanisms which may be involved are discussed.