Based on Theil index, the regional differences in carbon emissions of China’s logistics industry from 2015 to 2020 are explored. The three-stage SBM model and GML index model are constructed to analyze the efficiency from the static and dynamic perspectives respectively. Meanwhile, Moran ’I is used to reveal its temporal and spatial evolution characteristics. The results show that: there is a significant regional difference in carbon emissions from China’s logistics industry, showing a convergence characteristic of intra group convergence and inter group divergence. After adjustment, the efficiency is obviously improved, the eastern, northern and southern coastal areas have obvious efficiency advantages, while the northeast and southwest areas are in a disadvantageous position. Based on the analysis of external environmental factors, it is found that strengthening government support and leveraging location advantages are advantageous factors to reduce resource redundancy and promote efficiency. The total factor productivity shows an “M” type unstable fluctuation trend, and technological progress is the main factor affecting the total factor productivity. The total factor productivity of the southern coastal area, the middle reaches of the Yellow River, the north coastal area, the eastern coastal area, the southwest and the northwest areas is in a growing state, while the northeast area and the middle reaches of the Yangtze River are in a declining state. In terms of time series, the efficiency of China’s logistics industry has a significant spatial positive correlation, and has experienced an “inverse N” type evolution process of discreteness-agglomeration-discreteness. Spatially, it has basically formed a spatial agglomeration pattern of convergence in the eastern and northern coastal areas with high efficiency and convergence in the southwest and northeast areas with low efficiency.