The epistemological and methodological implications of complexity theory for understanding urban sprawl are discussed. It is argued that urban spatial forms, such as sprawl, emerge from nonlinear, self-organizational, and dynamic urban processes. Because of this, there cannot be a universal theory of sprawl and each case should be investigated within its context. The micro—macro problem provides the conceptual grounding for these investigations. Agent-based simulations can be used to investigate the micro—macro transformations in urban systems. Implications of complexity theory for understanding the role of urban policies are discussed.