Ventricular administration of the opioid dynorphin A(1-17) induces feeding in rats. Because its pharmacological characterization has not been fully identified, the present study examined whether a dose-response range of general and selective opioid antagonists as well as antisense oligodeoxynucleotide (AS ODN) opioid probes altered daytime feeding over a 4-h time course elicited by dynorphin. Dynorphin-induced feeding was significantly reduced by a wide range of doses (5-80 nmol i.c.v.) of the selective kappa(1)-opioid antagonist nor-binaltorphamine. Correspondingly, AS ODN probes directed against either exons 1 and 2, but not 3 of the kappa-opioid receptor clone (KOR-1) reduced dynorphin-induced feeding, whereas a missense oligodeoxynucleotide control probe was ineffective. Furthermore, AS ODN probes directed against either exons 1 or 2, but not 3 of the kappa(3)-like opioid receptor clone (KOR-3/ORL-1) also attenuated dynorphin-induced feeding. Although the selective mu-antagonist beta-funaltrexamine (20-80 nmol) reduced dynorphin-induced feeding, an AS ODN probe directed only against exon 1 of the mu-opioid receptor clone was transiently effective. Neither general (naltrexone, 80 nmol) nor delta (naltrindole, 80 nmol)-selective opioid antagonists were particularly effective in reducing dynorphin-induced feeding, and an AS ODN probe targeting the individual exons of the delta-opioid receptor clone failed to significantly reduce dynorphin-induced feeding. These converging antagonist and AS ODN data firmly implicate the kappa(1)-opioid receptor and the KOR-1 and KOR-3/ORL-1 opioid receptor genes in the mediation of dynorphin-induced feeding.