Global value chain (GVC) analysis draws on international political economy, development studies, economic geography and economic sociology to provide insight into the transnational organization of economic activities and how these interact with local actors and institutions to shape development processes. GVCs in specific industries tend to be ‘governed’, more or less explicitly, by identifiable sets of ‘lead firms’ that select suppliers, place orders, set requirements, and sometimes tightly coordinate the activities of suppliers and affiliated companies. This approach to GVC governance has provided rich insights into the emergence of spatially dispersed yet centrally coordinated production and distribution networks. As important as firm-level actors and their inter-relationships are, it is evident that actors, institutions, and norms external to the value chain also shape GVCs governance, for example through regulation, lobbying, civil society campaigns, and third-party standard setting. Institutional actors, including states and multilateral institutions shape GVCs by providing a mechanism for signatories to enforce, or not enforce, regulations and a platform for negotiating the terms of international trade agreements. Workers can also influence governance, especially when they are represented by labour unions with the ability to negotiate the terms of employment or call work stoppages at the level of the enterprise, industry, or broader economy. To provide an expanded analytical toolkit for this expanding field of inquiry, we offer a typology of power in GVCs along two dimensions: the arena of actors (dyads and collectives) and the precision of power (direct and diffuse). This yields four main forms of power: bargaining (dyadic and direct), demonstrative (dyadic and diffuse), institutional (collective and direct) and constitutive (collective and diffuse). This typology can help isolate various forms of power, and how they layer, evolve, consolidate and diffuse through distinct mechanisms and trajectories in view of better understanding how are GVCs governed, and with what distributional effects.