Abstract:Based on the data of urban innovation linkage capability of 17 Prefecture-level cities from 2014 to 2017 in Shandong Province,we use the methods of Improved Gravity Model construct Indirect weighting matrix for spatial association of urban innovation linkage capability. Furthermore,we use the methods of Social Network Analysis and QAP regression analyzing the spatial network structure characteristics and the formation mechanism of spatial conduction relationship of urban innovation linkage capability .The results show that:①The spatial density of urban innovation linkage capability from 2014 to 2017 in Shandong Province increased, and the overall connectivity of the network increased. The characteristics of the node structure indicate that the cities in the central and eastern parts of Shandong Province are at the center of the network, and they are easy to be spatially associated with other cities, and the degree of “bridge” that promotes spatial correlation between other cities is also higher. ② From the perspective of cyberspace characteristics, the “multi-triangle” in inland areas and the “inverted T-type” strong correlation network in coastal areas are expanding and the fringe cities rely on the development of core cities, which makes the whole network form a strong correlation and complex urban innovation linkage capability network. ③ The QAP regression and robustness test results show that the geographical distance, the difference of human capital level, the difference of policy support, and the difference of enterprise informatization level have a relatively stable impact on the spatial linkage network of urban innovation linkage capability in Shandong Province.