Abstract:The dual impact of R&D partners on innovation is more prominent in international contexts, but few studies deeply explore the link between international R&D partnerships and innovation success and failure. This paper scientifically measures the geographic proximity and geographic diversity of R&D partners, and applies a bivariate probit model to compare the impact of R&D partner selection on innovation success and failure based on 778 observations of 161 Chinese high-tech industrial enterprises from 2017-2021. It expands the application of economic geography and knowledge base theory in international R&D partner contexts. The results show that: (1) both geographic proximity and diversity of R&D partners positively affect innovation success and failure; (2) R&D partners with higher geographic proximity contribute more to innovation success; R&D partners with lower geographic proximity contribute more to innovation failure; (3) R&D partners with lower geographic diversity contribute more to innovation success; R&D partners with higher geographic diversity have a positive effect on both innovation success and failure, but the difference in effect is not significant.