A significant body of recent scholarship anticipates that global civil society (GCS) will eventually democratize international governmental institutions, reducing the intermediation of state actors and making international institutions more directly responsive to citizens. This paper argues that such optimism rests on flawed theorization and insufficient empirical analysis. Theories of GCS-driven democratization that envision GCS as an independent watchdog or deliberative, cosmopolitan space conflate political liberalism with functional democracy. They give no assurance that citizen-stakeholders are equally represented in GCS actions. This weakness is demonstrated in the case of the World Bank, where lobbying for policy changes during the 10th replenishment of funds for the Bank’s International Development Association (IDA) was dominated by a small group of privileged actors with strong ties to Bank donors. The paper dissects the incentives of the civil society organizations (CSOs) involved in the IDA and identifies the means by which the most influential CSOs achieved their objectives. Building on this analysis, it suggests a new theorization of global civil society that better accounts for competition between GCS actors and the enduring influence of the nation-state.