This paper proposes a model that can measure the R&D efficiency of each region (DMU) or each production unit while taking the inter-DMU competition and inter-subprocesses competition into account. The game cross-efficiency concept is introduced into the parallel DEA model. Furthermore, each DMU (subprocess) tries to maximize its own efficiency without harming the cross efficiency of each of the other DMUs (subprocess). We carry out an algorithm to obtain the best game cross-efficiency scores. This score has been proved to converge to a Nash equilibrium point. We use the proposed model to measure the R&D efficiency of the 30 provinces of China. The results show that the algorithm converges to a unique cross efficiency and our model indeed takes the bargaining power of DMUs and subprocesses into account.