Abstract The effect of prenatal alcohol exposure on serial-pattern learning was investigated in adult rats. Pregnant dams were administered a liquid diet which contained 35% ethanol-derived calories (35% EDC) on days 6–20 of gestation. Two control groups were included: a liquid diet control which was pair-fed and had sucrose substituted for ethanol (0% EDC); and a group fed standard lab chow throughout pregnancy. Offspring from each of these prenatal groups were trained to run a straight alley for food pellets. Pellet amount on successive trials was 14, 7, 3, 1, 7. Subjects were exposed to this pattern four times a day for 8 consecutive days. The mean latency to enter the goalbox for the four trials that comprised a monotonically decreasing series of pellet amounts (14, 7, 3, 1) was indirectly related to pellet amount. The mean latency on each trial and the degree of change in running speed across trials was found to be directly related to in utero alcohol exposure. Alcohol exposed offspring demonstrated less change in latencies over trials than controls. Also, the mean latency on the last trial for 7 pellets was shorter than that seen on the previous trial (1 pellet) and similar to that seen after the first trial in which 7 pellets were available. This indicated that satiation was not a factor in the increasing latencies. These data indicate a deficit in serial pattern performance as a function of prenatal alcohol exposure similar to what is observed in children with FAS.