Management researchers increasingly realize that some theories do not merely describe, but also shape social reality; a phenomenon known as "performativity." However, when theories become performative or even self-fulfilling is still poorly understood. Taking this gap in the research as our starting point, we develop a process model to show that new theories will only become self-fulfilling (1) if they motivate experimentation, (2) if experimentation produces anomalies, and (3) if these anomalies lead to a practice shift. On that basis, we identify six boundary conditions that determine whether theories will shape social reality. To illustrate our argument, we explore the conditions under which theories that postulate a positive link between corporate social performance and corporate financial performance may become self-fulfilling.