Performance on a Morris water task was examined in young rats whose mothers consumed a liquid diet consisting of 35% ethanol-derived calories (EDC) during pregnancy. Offspring of pair-fed (0% EDC) and ad lib lab chow (LC) dams served as controls. Rats were required to find a platform submerged below the surface in a pool of opaque water. A trial ended when the rat remained on the platform for 15 sec, or had been in the tank for 180 sec without reaching the platform. Subjects received 5 trials daily for 3 consecutive days, followed by reversal training on Day 4. Groups did not differ in swimming ability. On Day 1 there were no group differences among females in latency to reach the platform or in distance traveled, but male 35% EDC and 0% EDC animals had shorter latencies than LC controls. On Day 2, latencies and distance traveled of LC and 0% EDC controls decreased while 35% EDC animals showed no change from Day 1, so that alcohol-exposed rats took longer to reach the platform and traveled a greater distance than controls. On Day 3, 35% EDC females took longer than controls to reach the platform, and 35% EDC animals of both sexes traveled a greater distance than controls. Search patterns on the first reversal trial on Day 4 suggest the differences are in spatial processing and not learning per se, but more so in alcohol-exposed males than females. The impaired performance on this task suggests that prenatal alcohol exposure alters the ability to process spatial information.