Previous work has shown that opiate agonist infusion into the nucleus accumbens, a region implicated in reinforcement, stimulates food intake. In the present study, the effects of opiate antagonist infusion into this region were examined in two behavioral paradigms. In the feeding test, food-deprived animals were tested for intake of laboratory chow. In the sucrose intake test, sated animals familiar with a 20% sucrose solution were tested. Before these tests, the following drugs were bilaterally infused into the accumbens: naloxone (0, 1, 10 and 30 micrograms; equivalent to 2.8, 28 and 83 nmol, respectively), naltrexone (0, 0.2, 2 and 20 micrograms; 0.55, 5.5 and 55 nmol, respectively), beta-funaltrexamine (0 and 15 micrograms; 31 nmol), naloxonazine (0 and 10 micrograms; 15 nmol), naltrindole (trial 1: 0, 1, 10 and 20 micrograms; 2.2, 22 and 44 nmol, respectively; trial 2: 0, 0.1 and 0.5 micrograms; 0.22 and 1.1 nmol, respectively) and nor-binaltorphimine (0, 0.1, 1 and 10 micrograms; 0.14, 1.4 and 14 nmol, respectively). Naloxone and naltrexone both significantly reduced sucrose drinking and did not affect feeding. Naloxone infused into the dorsolateral striatum, as a control, had no effect on sucrose drinking. Accumbens infusion of the mu antagonist beta-funaltrexamine reduced both sucrose drinking and feeding. The mu 1 antagonist naloxonazine did not influence intake behaviors, with the exception of a decrease in duration of chow feeding. In contrast, the delta antagonist naltrindole markedly potentiated both sucrose drinking and duration of chow feeding. In a replication of this effect, systemic naltrexone given concurrently blocked the enhancement. The kappa antagonist nor-binaltorphimine did not influence any parameters of ingestive behavior. Although some treatments also decreased motor activities, the overall profile of behavior suggested specific effects on ingestive behavior. The putative contributions of mu and delta receptors within the nucleus accumbens to modulation of food reward are discussed.