Collaboration between actors in political decision-making processes is crucial from both an actor and a process perspective. Previous studies have highlighted the role of preference similarity, power, and opportunity structures as drivers of collaboration. However, these studies have focused on single policy sectors and have therefore overlooked possible differences in effects across sectors, as well as interactions between sectors. This article innovates by taking a cross-sector perspective. Applying exponential random graph models to collaboration networks covering 11 decisionmaking processes, we show that relational opportunity structures have a fairly consistent influence on collaboration, whereas the effects of social and institutional opportunity structures vary across processes. The effect of institutional opportunity structures is contingent on the importance of institutional arenas. Our hypothesis on cross-sector influences receives partial support. Opportunity structures and interactions between sectors add to preference similarity and power, which have a strong and robust influence on collaborative tie formation.