A case might be made that we have not learned very much about leadership in complex organizations. One factor contributing to this lack might be the narrowness of the traditional approach to leadership. By defining leadership as an interpersonal influence process, a research literature has been generated that focuses almost exclusively on leader-subordinate relationships, two or so styles, and group outcomes. Such a focus does not ask questions about the environmental and organizational context of leadership and the impacts it may have. A second factor contributing to narrowness might be an overdependence on one methodology for studying leadership, namely the survey. One might argue that direct observation, or at least multiple methodologies, could be usefully applied to understanding leadership. It was with these conceptual and methodological concerns in mind that a simulation was designed for use in leadership research. Its goals were both simple and complex: to mirror as realistically as possible the demands of a typical managerial job, to have real managers run the simulated company however they chose, and to bring a full methodological arsenal to bear on learning something new about leadership. By watching managers cope with the problems and challenges of organizational life, we hoped to come up with some better questions to guide future research. This paper uses Looking Glass, Inc. (a person-centered simulation of management in a complex organization) as an example of how simulation can be used to examine research questions that are designed-in, phenomena that occur because the complexity of real organizations was preserved, and possible research interventions. The nature of this simulation also permits the use of an unusual number of research methods, and these are outlined. The paper concludes by addressing three major questions of the simulation: Is it valid? Is it practical? Is it a game?