Urban agglomerations are areas of high energy consumption and carbon emissions, and are also the key areas where green development can be realized by reducing pollution and carbon emissions. However, the regional differences and evolutionary trends of carbon emission intensity in urban agglomerations have not been adequately studied.Taking three major city clusters in China as examples, this paper uses Dagum Gini coefficient and its decomposition method, Kernel density estimation method, Markov chain and spatial Markov chain to study the regional differences, dynamic evolution characteristics and spatial spillover effects of carbon emission intensity of urban agglomerations, and accordingly proposes differentiated emission reduction and carbon reduction policies. The following results were obtained: (1) The overall carbon emission intensity of the three major city clusters and each individual urban agglomeration were found to have declined significantly over time, with an overall spatial pattern of "high in the north and low in the south", with inter-group differences being the main source of the overall differences. (2) The imbalance in carbon emission intensity between cities was more obvious within the Beijing-Tianjin-Hebei urban agglomeration, while the synergistic emission reduction effect of the Yangtze River Delta and Pearl River Delta urban agglomerations increased over the study period. (3)The probability of a city maintaining a stable level of carbon emission intensity was Feng Dong