A wide variety of drugs used in medicine and on a non-prescription basis have abuse liability. However, not all psychoactive drugs have demonstrable reinforcing properties either in animal models or in humans. In order to predict the abuse liability of a compound or to understand determinants of the abuse liability, several animal testing procedures have been developed. These procedures include tests for physical dependence, tolerance, disruption of ongoing behaviors, discriminative stimulus properties, and the direct or indirect assessment of the reinforcement. While tests for the production of tolerance and physical dependence may not predict abuse liability, they would significantly modify the abuse risk. This paper will focus on the methods available for the assessment of the reinforcing properties and the advantages of using genetically defined mice to better understand the determinants of the reinforcing properties and to isolate phenotypes that are hyper-responsive and hypo-responsive to the reinforcing properties of drugs under study. Examples of genetically determined differences are provided and areas of inadequate information are discussed.