Adult female Sprague-Dawley rats were prepared with permanent cortical EEG and temporalis EMG electrodes and i.v. cannulae. They were made tolerant to and physically dependent on morphine by automatic, hourly injections. These physically dependent rats were then trained to lever press for 10 mg/kg injections of morphine on a fixed ratio (FR) schedule of reinforcement. Upon stabilization of morphine self-administration at a FR-10, dynorphin-[1-13] (DYN) or D-ala2-dynorphin-[1-11] (D-ala2-DYN) at doses of 125 or 250 micrograms/kg/inj was substituted for morphine. Rats self-administered these opioid-like peptides at both dose levels. As expected, self-injections were more numerous at the lower dose. No signs of morphine withdrawal were seen during the peptide substitutions. Following DYN or D-ala2-DYN abstinence, no withdrawal symptoms were noted. The question is raised as to whether DYN or D-ala2-DYN and morphine are producing their reinforcing effects in sustaining self-administration via the same receptor populations. Since morphine abstinence is associated with severe withdrawal symptoms and the peptides studied are not, the involvement of separate receptor populations in the process of dependence on morphine and these opioid-like peptides is indicated. In conclusion, both a mu and two kappa agonists exhibited an analogous reinforcing property in the rat. However, the degree of physical dependence and the intensity of withdrawal differed; being higher with the mu agonist and lower with the kappa agonists.