The use of illicit drugs by American youth rose dramatically during the 1990s. Reducing these trends is an important policy objective. However, for policies to be effective it is important to understand the key causal links that lead to substance use and abuse. Policy makers must understand whether attempts to reduce the demand for one drug have impacts on the current and future use of other drugs. This paper overviews an economic approach to modeling drug use, addiction and gateway effects, emphasizing the potential of this method for identifying causal links in consumption. The paper demonstrates how this multi‐substance theory of drug use leads to empirical specifications that can identify the impact of consumption of one drug on the contemporaneous and future consumption of other drugs. This is followed by a discussion of what types of data would be needed to estimate these effects.