Recent studies have shown that post-retrieval extinction training attenuates fear and reward-related memories in both humans and rodents. This noninvasive, behavioral approach has the potential to be used in clinical settings to treat maladaptive memories that underlie several psychiatric disorders, including drug addiction. However, few studies to date have used a post-retrieval extinction approach to attenuate addiction-related memories. In the current study, we attempted to disrupt cocaine-related memories by using the post-retrieval extinction paradigm in male Sprague Dawley rats. Results revealed that starting extinction training 1 h after cocaine contextual memory was retrieved significantly attenuated cocaine-primed reinstatement of conditioned place preference (CPP) and relapse of cocaine CPP (drug-free and cocaine-primed) following 30 days of abstinence. However, animals that did not retrieve the contextual cocaine memory before extinction training, or animals that began extinction training 24 h after retrieval (outside of the reconsolidation window), demonstrated normal cocaine CPP. Conversely, animals that received additional CPP conditioning, rather than extinction training, 1 h after reactivation of cocaine memory showed enhanced cocaine CPP compared with animals that did not reactivate the cocaine memory before conditioning. These results reveal that a behavioral manipulation that takes advantage of reconsolidation and extinction of drug memories may be useful in decreasing preference for, and abuse of, cocaine.