Simulation has been adopted in many disciplines as a means for understanding the behavior of a system by imitating it through an artificial object that exhibits a nearly identical behavior. Although simulation approaches have been widely adopted for theory building in disciplines such as engineering, computer science, management, and social sciences, their potential in the IS field is often overlooked. The aim of this paper is to understand how different simulation approaches are used in IS research, thereby providing insights and methodological recommendations for future studies. A literature review of simulation studies published in top-tier IS journals leads to the definition of three classes of simulations, namely the self-organizing, the elementary, and the situated. A set of stylized facts is identified for characterizing the ways in which the premise, the inference, and the contribution are presented in IS simulation studies. As a result, this study provides guidance to future simulation researchers in designing and presenting findings.