Despite numerous criticisms regarding its applicability and utility, the traditional cost-of-illness approach has remained the primary method for evaluating and measuring the economic cost of substance abuse. We provide motivation for the field to consider an alternative theoretical construct for measuring the economic cost of drug abuse that is more closely embedded within the tradition of Economics, considering both the short and long term costs of abuse, the timing of these costs, and who bears these costs. We describe work currently underway at the RAND Corporation, which applies this alternative framework to assess the economic cost of drug abuse in the United States using secondary data sources and simulation methods. At the heart of the exercise is a microsimulation model of the epidemiology of substance use and the impact of this use on other life events (drug treatment, crime, and productivity). We demonstrate the methodological strengths of using a microsimulation model to describe drug use in the population over other alternative modeling strategies. Our beta model simulates for a group of individuals who start the world at age 12 their initiation, escalation, and possible dependence on illicit drugs, evaluates the impact of treatment for these individuals, and the interaction with the criminal justice system. The results provide justification for re-thinking how policy makers consider the burden of drug use on society and show how using this new paradigm could improve policy decisions.