This study explores the configuration of multi-partner collaborations (MPCs) and the inherent trade-offs these configurations present. While the pooling of resources and sharing of risks among multiple partners offer substantial benefits, they also introduce significant governance complexity and collaboration costs associated with multilateral interactions. Previous research suggests that collaboration challenges may escalate more rapidly than resource benefits, setting a limit on the size and diversity of these collaborations. Among others, dominant-partner governance has been proposed as a solution, where powerful partners act as peacekeepers and coordinators to mitigate collaboration challenges, thereby supporting larger and more diverse groups. This solution, however, assumes that the dominant partner possesses both the expertise and objectives aligned with the collective entity, an assumption our research critically reevaluates. We examine the effects of dominant-partner governance within the context of project-financed infrastructure projects. We specifically investigate the role of government as a dominant partner and find that while such involvement tends to reduce the size and diversity of expertise within the group, these negative impacts are moderated by the government’s collaborative capabilities and its reliance on resources from private partners. Through an analysis of an original dataset covering 1,233 infrastructure projects across 68 countries and five sectors from 1997 to 2020, our findings contribute to a nuanced understanding of MPC dynamics and configuration.