Applying the panel data of 16 districts in Beijing, China from 2009 to 2020 as the research object, this study measures and analyzes the carbon emission intensity and the level of industrial structure upgrading. Based on the above results, a spatial econometric model is established to analyze the spatial spillover effect of industrial structure upgrading on carbon emission intensity. Conclusions are drawn as follows: (a) In 2009, 2015 and 2020, the carbon emission intensity in most districts of Beijing has decreased, and in some areas even decreased significantly. The upgrading of industrial structure in all districts has been improved (b). According to the results of spatial autocorrelation, the carbon emission intensity in Beijing shows significant positive spatial autocorrelation in 2009 and 2020, while negative spatial autocorrelation in 2015; The upgrading of industrial structure in Beijing shows significant positive spatial autocorrelation in 2009, 2015 and 2020 (c). The regression results of the spatial econometric model show that industrial structure upgrading not only reduces the carbon emission intensity of the region, but also decreases the carbon emission intensity of the surrounding areas.