OBJECTIVE: To present a novel network-based framework for the study of collaboration in surgery and demonstrate how this can be used in practice to help build and nurture collaborations that foster innovation. BACKGROUND: Surgical innovation is a social process that originates from complex interactions among diverse participants. This has led to the emergence of numerous surgical collaboration networks. What is still needed is a rigorous investigation of these networks and of the relative benefits of various collaboration structures for research and innovation. METHODS: Network analysis of the real-world innovation network in robotic surgery. Hierarchical mixed-effect models were estimated to assess associations between network measures, research impact and innovation, controlling for the geographical diversity of collaborators, institutional categories, and whether collaborators belonged to industry or academia. RESULTS: The network comprised of 1700 organizations and 6000 links. The ability to reach many others along few steps in the network (closeness centrality), forging a geographically diverse international profile (network entropy), and collaboration with industry were all shown to be positively associated with research impact and innovation. Closed structures (clustering coefficient), in which collaborators also collaborate with each other, were found to have a negative association with innovation (P < 0.05 for all associations). CONCLUSIONS: In the era of global surgery and increasing complexity of surgical innovation, this study highlights the importance of establishing open networks spanning geographical boundaries. Network analysis offers a valuable framework for assisting surgeons in their efforts to forge and sustain collaborations with the highest potential of maximizing innovation and patient care.