Selective cross-tolerance is often used as evidence to differentiate opioid receptor subtypes. We used this strategy to study operant behavioral effects of the opioid peptides, [D-Ala2, NMePhe4, Gly-ol5]enkephalin (DAGO), [D-Pen2.5]-enkephalin (DPDPE) and dynorphin, agonists highly selective for mu, delta, and kappa receptors, respectively. Food-deprived rats were trained to lever-press on a fixed-interval 3-min schedule of food-reinforcement. Time-effect and dose-effect curves were generated for each of the peptides, as well as for morphine, administered ICV, 5 min prior to the 1-h operant session. Experiments were performed on untreated subjects and on subjects receiving chronic infusion of morphine (10 mg/kg/day) from osmotic pumps. In untreated animals, morphine and the mu-selective peptide, DAGO, induced relatively long-lasting dose-related decreases in responding, whereas the non-mu agonists, DPDPE and dynorphin, induced only transient effects: response rates increased at low doses and decreased at high doses. Animals receiving chronic morphine infusion were tolerant to the rate-decreasing effects of morphine and DAGO; ED50s increased by factors of 8 and 6, respectively. There was some evidence of cross-tolerance to DPDPE and of sensitization to dynorphin in the morphine-maintained animals.