Our study contributes to the nascent discourse on enterprise architecture management (EAM) benefits by proposing a theory-led, empirically validated model to explain how EAM benefits unfold. Drawing on the resource-based theory, and based on empirical insights from 8 case studies, we find that EAM does not create benefits per se, but only creates value if an organization develops four second-order EAM capabilities – EA modeling, EA planning, EA implementation, and EA governance. The discovery that EAM resources only unfold their potential when forming EAM capabilities casts doubt on the established practice of initiating EAM as a modeling and documentation endeavor.