AbstractThe value of craving as a construct in the substance use disorders literature stems from its purported ability to predict actual alcohol or psychoactive drug consumption. Empirical reports of craving's relationship to actual use, however, are quite mixed. It is argued that craving's relationship to use variables such as relapse will remain unclear until individual differences related to acting on craving are considered. Such potential mediators of craving's relationship to relapse, such as impulsivity, are evaluated. The implications of this argument for the construct of craving are discussed.