Abstract: Based on Chinese economic census data, the geographic weighted regression model is used to study the influence of regional innovation source factors on regional innovation performance and its spatial heterogeneity. The results show that the most important source of regional innovation in China is indigenous R&D. Among other sources of innovation, domestic technology transfer and absorption also have positive effects on regional innovation. Effects of international technology transfer, export and FDI are generally negative. In different provinces, the impacts of innovation factors on the regional innovation performance are spatially heterogeneous. The results of geographic weighted regression reveal that the spatial characteristics of the influence of innovation factors on regional innovation generally demonstrate gradient transfer from the eastern developed regions to the western regions. In the eastern and central regions, indigenous R& D has become a major source of innovation, the role of which on regional innovation is more favorable and significant, while in the western region, innovation sources except for indigenous R& D, domestic technology transfer and absorption still play an important role. Geographic weighted regression method is proper tool to identify the spatial heterogeneity of the effects.