We argue that inadequate frameworks to compare similar gang control strategies, and the scarcity of well-designed evaluations have hindered our ability to determine the effectiveness of existing programs. This article proposes a new typology of gang control strategies to use with logic models as tools to improve gang program evaluation.