Recent work has shown that transport of food items from open, exposed food sources to a covered shelter is reduced by drugs thought to have anxiolytic properties in rodents and humans. We studied the effects of amphetamine and the dopamine D2/3-receptor antagonist, raclopride, in this test of food transport that pits immediate food consumption against exposure in an open space. Rats traveled from a home cage along an elevated beam to obtain single food items of varying sizes located at one of 12 distances from the home cage. Large food items and items located close to the home cage were carried back and consumed inside the cage. Small items and items located farther from the cage were eaten immediately at the food source while sitting on the beam. Amphetamine sulfate (0.001–2.0 mg/kg, i.p.) decreased eating on the beam and increased carrying of food items to the home cage. Raclopride (0.005–0.2 mg/kg, i.p.) tended to reduce carrying of food to the home cage, but 0.05 mg/kg raclopride did not block the increase in food carrying seen with amphetamine treatment (2 mg/kg). The increased food carrying seen with amphetamine is opposite to the effect produced by anxiolytic drugs, raising the possibility that amphetamine promotes carrying by increasing defense or ‘anxiety’. Consistent with this hypothesis, amphetamine (2 mg/kg; the maximally effective dose in the food-carrying experiment) decreased open-arm exploration in the elevated plus-maze, considered to be an anxiogenic effect. These results indicate that stimulation of monoaminergic neurotransmission increases food transport from exposed food sources to a shelter; D2/3-receptor blockade tends to reduce it. The food-carrying test provides a rich, ethologically valid paradigm to assess the effects of psychoactive drugs on species-specific, defensive behaviors in rodents.