Knowledge spillovers are a central part of knowledge accumulation. The paper focuses on spillovers that occur through the interaction between different researchers or developers that collaborate in different research projects. The paper distinguishes between project spillovers and contributors' spillovers and between direct and indirect spillovers. The paper constructs a unique data set of open source software projects. The data identifies the contributors that work in each project and thus enable us to construct a two-mode network: a Project network and a Contributor network. The paper demonstrates that the structure of these networks is associated with project success and that there is a positive association between project closeness centrality and project success. This suggests the existence of both direct and indirect project knowledge spillovers. We find no evidence for any association between contributor closeness centrality and project success, suggesting that contributor spillovers play a lesser role in project success.