Abstract: From the perspective of knowledge space diffusion, the paper makes a theoretical analysis of the spatial correlation of industrial technological innovation efficiency. The paper uses the window MinDS super-efficiency model to measure the industrial technology innovation efficiency of 30 provinces from 2009 to 2015 in China and determines the spatial correlation of industrial technology innovation efficiency between provinces with the adjusted gravity model. Then the paper uses the method of social network analysis to reveal the characteristics and the determinants of spatial correlation network. The results show that there exists a more significant and complex network structure in provincial industrial technology innovation efficiency. The spatial correlation is getting closer and there is no hierarchical in the network. Jiangsu, Guangdong, Shandong and other Eastern and central developed provinces are at the core of the network, while Heilongjiang, Jilin, Guangxi and other western provinces are at the edge of the network. There are more intra-block correlations than inter-block correlations in the network. Spatial proximity, convenient transportation, similarity of industrial structure and difference of industrial technology innovation level have significant positive effects on spatial correlation of the provincial industrial technology innovation efficiency, while difference of informatization level has significant negative effects on spatial correlation. The study provides policy reference for the establishment of cross-regional synergistic promotion mechanism of industrial technology innovation efficiency and the implementation of regional innovation driving strategy.