Researchers in many fields have adopted simulation to understand a system’s behavior by imitating it through an artificial object that exhibits nearly identical behavior. Although simulation approaches have been widely adopted for theory building in fields such as engineering, computer science, management, and social sciences, researchers in the IS field often overlook their potential. In this paper, we examine how IS research uses different simulation approaches and, thereby, provide insights and methodological recommendations for future studies. From reviewing the literature on simulation studies published in top-tier IS journals, we define three classes of simulations: the self-organizing, the elementary, and the situated. We identify a set of stylized facts for characterizing the ways in which IS simulation studies present the premise, the inference, and the contribution. As a result, we provide guidance to future simulation researchers in designing and presenting their findings.