Reinstatement of previously extinguished instrumental responding for drug-related cues has been used as an animal model for relapse of drug abuse, and is differentially affected by inactivation of the core and shell subregions of the nucleus accumbens (NAc). To compare the roles of these subregions in reinstatement induced by cues associated with natural and drug rewards, the present study assessed the effects of inactivation of the NAc core and shell on cue-induced reinstatement of food-seeking behavior. Rats acquired a lever pressing response for food reward paired with a light/tone conditioned stimulus (CS). They were then subjected to extinction, where both food and the CS were withheld. Reinstatement of responding was measured during response-contingent presentations of the CS. Following saline infusions into the NAc core or shell, rats displayed a significant increase in lever pressing during reinstatement sessions. Inactivation of the core, induced by infusion of GABA agonists muscimol and baclofen, attenuated responding for the CS, but did not affect pavlovian approach toward the food receptacle. In contrast, inactivation of the shell had the opposite effect, potentiating responding relative to vehicle treatments. These data suggest that the NAc core and shell play opposing, yet complementary roles in mediating the influence that food-associated conditioned stimuli exert over behavior. The core enables reward-related stimuli to bias the direction and vigor of instrumental responding. In contrast, the shell facilitates alterations in behavior in response to changes in the incentive value of conditioned stimuli. The fact that the NAc core appears to play a similar role in cue-induced reinstatement induced by both natural and drug rewards suggests that this region of the ventral striatum may be a final common pathway through which both drug- and food-associated stimuli may influence the direction and magnitude of ongoing behavior.