In formulating policies for drug offenders, lawmakers must decide concrete questions about such matters as legal jurisdiction, burdens of proof, and reporting of progress information. Although these decisions may seem incidental to treatment and beyond the purview of science, they are based on empirically testable assumptions about the behavior of drug abusers and have a direct bearing on the efficacy of drug treatment interventions. Unfortunately, these assumptions have generally not been subjected to empirical inquiry. As a result, drug policy continues to be crafted by non-scientific advocates and subjected to popular vote by an insufficiently informed public. This article identifies several empirically answerable questions that underlie critical decision points in criminal statutes for drug offenders, reviews the available research evidence relevant to these questions, and encourages drug abuse researchers to conduct studies aimed squarely at informing these policy-relevant decisions. © 2003 Elsevier Inc. All rights reserved.