Conditioned Place Preference (CPP) is one of the models most frequently used to study drug use and addiction in animals such as mice and rats. CPP has a number of advantages, including low cost, ease of use and versatility. However, the model also has disadvantages: the novelty confound, difficulty in creating dose-effect curves, and inability to study discrete cues. It is important to determine whether alternate classical conditioning models can eliminate some of the limitations of CPP while maintaining its advantages. Here we did not find conditioned preference to odors. However, we were able to both demonstrate and later eliminate taste aversion. We found that mice did not spend more time investigating a scent after it had been repeatedly paired with morphine. Mice did develop an aversion for flavored food after it had been paired with morphine, which is consistent with previous research. Most importantly, when mice were given repeated injections of morphine to postpone withdrawal, they no longer developed an aversion for flavored food that was paired with morphine. Our results demonstrate that it may be possible to eliminate aversion and possibly even to create a preference for food paired with a drug, despite the fact that previous research has almost exclusively shown taste aversion. Demonstrating the practicability of eliminating taste aversion is the first step in establishing CTP as a viable model for drug abuse. Although more research is needed to further develop the model, CTP could serve as a useful complement or even replacement for CPP.