AbstractThis paper contributes to the development of better methods for the conceptualization of system dynamics models. It does so by looking at the model building process as a theory building process and by reviewing theory building approaches in other social sciences, such as grounded theory and case study research. Concepts of these modes of qualitative research constitute a tool set to build relevant system dynamics models, grounded in data, and with higher potential to provide rigorous and relevant generic structures. The deductive capability of simulation models, on the other hand, helps to close the loop in the theory building process, putting together the inductive and deductive stages in theory building. Copyright © 2008 John Wiley & Sons, Ltd.