Biotechnology industry is technology-intensive, highly innovative and high value-added industries. Especially in the pharmaceutical industry. Because biotech products subject to regulatory control. Resulting in long product development, funds raised is not easy because high investment risk. Difficult for many businesses to support to the stage to market, so efficient R&D performance is important foundation of the development of the biotechnology pharmaceutical industry. The other hand, the U.S. biotech industry become the industry leader because a steady stream of innovative technologies from academic and research institutions spread to the industry. The formation of social and technological interwoven network relationships in Scientists and industry. Therefore, this study applied the concept of Actor-Network Theory(ANT), the formation of a heterogeneous network in actor interessment, knowledge translation, situational conversion. Explore how to improve the biotechnology industries’ R&D performance by network cooperation. Although researchers have used multi-stage DEA and variables to measure R&D performance in the past, but they ignores the network concept into R&D performance evaluation studies. Therefore, this study establishes R&D performance model by situational conversion of biotechnology industry in Actor-Network Theory and network DEA -technology creation capability, technology diffusion capability and value creation capability, and analyze R&D performance of 22 biotechnology pharmaceutical corporations in Taiwan. The study finds the overall Taiwanese pharmaceutical industries’ R&D performance should be improved. Technology creation capability shows insufficient domestic patent. The technical cooperation group is more efficient than independent development in benchmarking. It shows R&D performance can be improved by Network cooperation. This study suggests that biotechnology industries create obligatory passage point(OPP) in ANT by industry-academic-government heterogeneous network collaboration and industry clusters. specialization and division of labor establish value chain can not only reduce the risk of operation and R&D but also step up industry scale and R&D performance.
