Collaboration networks are the main mechanisms through which regional innovation capacities are enhanced. These networks have been analyzed by using data either on research projects or patents. However, analyzing only one type of network can limit our understanding of regional innovation dynamics. In this paper, we investigate the drivers of network evolution in two inter-regional network types within the European Union: R&D project networks and co-patenting networks. Specifically, we examine data on Framework Program project collaborations and patent collaborations at the NUTS2 regional level between 2011 and 2019. Our empirical strategy is based on Stochastic Actor-Oriented Model where we analyze and compare the factors that affect the evolution of two networks. We find that path-dependent processes resulting from past networks, such as triadic closure, preferential attachment, and tie repetition, play a more important role in explaining the evolution of R&D project networks compared with patent networks. On the other hand, regional characteristics, particularly the patenting intensity of a region and similarity in patenting between regions appear to be more influential in shaping patent networks. These results have important implications for the future evolution of knowledge networks in Europe.