The growth of global public-private partnerships for health has opened up new spaces for civil society participation in global health governance. Such participation is often justified by the claim that civil society organizations, because of their independence and links to communities, can help address democratic deficits in global-level decision-making processes. This article examines the notion of ‘civil society engagement’ within major public–private partnerships for health, where civil society is often said to play a particularly important role in mediating between public and private spheres. How do major global health partnerships actually define ‘civil society’, who represents civil society within their global-level decision-making bodies, and what formal power do civil society representatives hold relative to other public and private-sector partners? Based on a structured analysis of publicly available documents of 18 of the largest global public–private partnerships for health, we show that many of them make laudatory claims about the value of their ‘civil society engagement’. Most use the term ‘civil society’ to refer to non-governmental organizations and communities affected by particular health issues, and state that they expect these actors to represent the needs and interests of specific populations in global-level decisions about strategies, funding models and policies. Yet, such civil society actors have a relatively low level of representation within the partnerships’ boards and steering committees, especially compared with private-sector actors (10.3 vs 23.7%). Moreover, there is little evidence of civil society representatives’ direct and substantial influence within the partnerships’ global-level governing bodies, where many decisions affecting country-level programmes are made. Rather, their main role within these partnerships seems to be to implement projects and advocate and raise funds, despite common discourses that emphasise civil society's watchdog function and transformative power. The findings suggest the need for in-depth research into the formal and informal power of civil society within global health governance processes.