Drug-paired conditioned reinforcers can maintain persistent instrumental responding, thus providing a model of some aspects of long-term drug addiction. The purpose of the present study was to investigate the effects of inactivating the dorsal striatum (DStr), nucleus accumbens (NAcc) core, or NAcc shell on different types of responding, each maintained by drug-paired conditioned reinforcers. Inactivations were achieved by infusing a combination of baclofen and muscimol prior to (1) persistent responding for a drug-paired conditioned reinforcer, (2) reacquisition of this instrumental response after extinction by omission of the contingent conditioned stimulus (CS), or (3) CS (cue)-induced reinstatement of the original (and different) instrumental response that had previously delivered cocaine. Inactivation of the DStr attenuated persistent responding for a cocaine-paired conditioned reinforcer, as well as its reacquisition after extinction of this response, while the only effect of inactivation of the NAcc shell was to increase CS (cue)-induced reinstatement of the extinguished instrumental response that had previously delivered cocaine. Inactivation of the NAcc core affected all measures of responding maintained by drug-paired conditioned reinforcers. These results are discussed with reference to the neural systems involved in different aspects of responding maintained by drug-paired conditioned reinforcers.