The purpose of this study was to understand the capacity for diffusion of computer science instruction in an urban school district. Two types of networks, general instructional support and computer science-related support, were described and investigated. Social Network Analysis was used as the primary method to examine the structure of and relationships between the networks. Results suggest that even in schools with dense and distributed instructional support networks, sparse and centralized systems of ties are characteristic of DLCS support networks. Further, an analysis of networks with and without team-supported ties indicates that formal structures for collaborative teaming are critical sources of social capital for teachers and are essential for the diffusion of high quality DLCS instructional practices. Multinomial logistic regression indicated a significant positive relationship between teachers’ self-efficacy and in-degree centrality, and a significant but negative relationship between seniority and out-degree.