Abstract：
The essay uses spatial autocorrelation Moran’I, spatial error model and spatial lag model of spatial econometrics, combined with improved Romer-Jones Knowledge Production Function(KPF),to analyze spatial knowledge spillover effects of china. Analysis results show that R&D input, R&D personnel, stock of knowledge, and foreign introduced technology have the positive contribution for the production of new knowledge, the contribution degree of various factors is different, among them, the average contribution rate of the stock of knowledge is the biggest (0.460).The production of new knowledge of China has provincial spatial spillover effect. With new knowledge production of surrounding areas an average growth of 1%, the region’s new knowledge production will increase by an average of 0.469%.
2016 
