Developing low-carbon logistics requires an understanding of the driving factors for carbon emissions. We employed the Comparative Study on Urban Transport and the Environment (CUTE) framework to identify the driving factors of China’s logistics carbon emissions. Then, the Generalized Fisher Index Decomposition (GFID) model was adopted to decompose the effects of the driving factors. Finally, we tracked the spatial dynamics trajectory of each driving effect based on the gravity model. Our results showed that technical intensity and transport structure promoted carbon emissions, while technical efficiency and agglomeration curbed carbon emissions. The gravity centers of transport structure and technical efficiency converged with that of carbon emissions, whereas the gravity centers of technical intensity and industry agglomeration diverged from that of carbon emissions. The driving effects showed an obvious spatial heterogeneity, which indicated that carbon reduction policies should be formulated according to the local situation.