Abstract:This paper constructs the empirical agent variables of the agglomeration scale of manufacturing industry and producer service industry based on the location entropy model, Then we use the expanded Griliches-Jaffe knowledge production function , restructured and logarithmic-ally processed to construct a Log-linear model of the impact of industrial agglomeration on regional innovation efficiency, Using macroeconomic panel data from Beijing, Shanghai, Guangzhou and Shenzhen from 2010 to 2020, this paper empirically analyzes the agglomeration effect and crowding effect of industrial agglomeration on the efficiency of scientific and technological innovation in China"s four first-tier cities, the results show:(1)There is an inverted U-shaped non-linear relationship between the industrial agglomeration scale and the development of sci-tech innovation efficiency in the four first-tier cities of Beijing, Shanghai, Guangzhou and Shenzhen;(2)There is a linear relationship between the agglomeration scale of producer services and the development of science and technology innovation efficiency in the four first-tier cities of Beijing, Shanghai, Guangzhou and Shenzhen;(3)There is also a “Inverted U-shaped” nonlinear relationship between the inter-industry cluster of manufacturing and producer services in the four first-tier cities of Beijing, Shanghai, Guangzhou and Shenzhen, and the development of their sci-tech innovation efficiency, finally, we give some policy suggestions.