The rise in the recreational use of cocaine in the last two decades has resulted in a growing health concern about fetal drug exposure. In exposed children, investigators have noted altered cognitive performance in complex or distracting, but not more controlled, situations. In rodent models, deficits in short-term memory have been noted in some, but not all, paradigms, although these studies also differ in routes of administration and dosing models. Here, we report short-term memory deficits in rats prenatally exposed to cocaine using an intravenous administration model to closer mimic cocaine doses and pharmacokinetics seen with human use. A spontaneous two object recognition task was used to avoid (1) clearly aversive or rewarding components, (2) reference memory component and (3) the use of external motivators, such as swimming stress or food deprivation/rewards. In this task, adolescent and adult male rats exposed to cocaine in utero demonstrated deficits in short-term memory compared with saline controls. No difference in the time spent exploring the objects or the number of failures was noted between the prenatal cocaine and saline rats. This study suggests that prenatal exposure to cocaine can result in a long-lasting deficit in non-spatial, short-term memory in a spontaneously performed task.