Abstract: Carbon emission regulation should not only focus on its own emission reduction, but also examine the inter-provincial carbon emission linkage. To clarify the network characteristics and influencing factors of spatial correlation of carbon emission at the national level is the basic premise of overall carbon emission control and sustainable development. Based on the panel data of China's provinces from 2015 to 2019, social network analysis (SNA) was used to conduct empirical analysis on the spatial correlation of carbon emissions, and the influencing factors were explored by QAP method. The results show that China's carbon emissions show typical spatial network characteristics, and the inter-provincial carbon emissions are increasingly correlated, and the stability and complexity of the network are gradually improved. Shanghai, Jiangsu, Beijing and other places live in the network center, and play the role of "bridge"; The economically developed eastern coastal region plays the role of "two-way spillover" and "net benefit" in the network, while the central and western inland regions with rich energy reserves belong to the "broker" and "net spillover" plate. Differences in geographical proximity, industrial structure, income, technological innovation and population density have a significant positive impact on the spatial correlation of carbon emissions. The more similar the energy consumption is, the stronger the correlation of carbon emissions is. Therefore, in carbon emission governance, it is necessary to base on the concept of regional collaborative governance, to establish a "leader-follow" carbon emission reduction mechanism, to give full play to the strategic advantage of "a national chess game", to formulate differentiated carbon emission reduction policies, and to promote the formation of a balanced carbon emission reduction pattern at the national level.