Abstract: As an important indicator to measure energy conservation and emission reduction, carbon emission intensity plays a vital role in promoting the transition to a green and low-carbon economy. Based on the panel data of 30 provinces from 2000 to 2021, and by using exploratory spatial data analysis, kernel density estimation, standard deviation ellipse analysis and geodetector, this paper comprehensively analyzed the spatiotemporal dynamic evolution characteristics, spatial agglomeration effect and multi-factor driving mechanism of carbon emission intensity in China. The results show that during 2000 to 2021, the inter-provincial carbon emission intensity showed a downward trend of "first slow and then rapid", and the time scale decreased from 38.464 t·(10⁴ yuan)^-1(2000) to 12.872 t·(10⁴ yuan)^-1(2021), meanwhile the carbon emission intensity of all regions also showed a downward trend. On a spatial scale the western region exhibited the highest carbon emission intensity, followed by the northeastern region and the central region, while the eastern region had the lowest carbon emission intensity. The kernel density curve of carbon emission intensity shows an overall leftward shift, with the main peak becoming taller and narrower, and exhibiting a multi-peak distribution, indicating that the carbon emission intensity is in a dynamic decreasing trend, and the absolute difference in carbon emission intensity among provinces is narrowing. Carbon emission intensity shows the spatial distribution characteristics of northeast-southwest and the center of gravity transfer direction is southwest-northeast-northwest-north, mainly concentrated in Shanxi Province and Shaanxi Province. The spatial pattern of carbon emission intensity is high in the north and low in the south, with significant spatial dependence and increasing agglomeration; The hot spot analysis shows an agglomeration effect of "southeast cold-northwest heat". Among the driving factors of carbon emission intensity, the level of foreign investment and structure of energy consumption have the most significant impact on carbon emission intensity, moreover, the interactive effects of multiple factors are greater than those of single factors. Efforts for reducing carbon emissions should comprehensively consider the multi-factor driving mechanism. The research results indicate that it is crucial to formulate and implement regional low-carbon strategies. Proposing targeted carbon reduction measures based on the influence levels of driving factors can provide valuable references for achieving the dual carbon goals of China.