Improving urban ecological efficiency is an important aspect of high-quality development of the Yangtze River Economic Belt. In this study, the minimum distance to strong efficient frontier data envelopment analysis (MinDS) model was used to measure the ecological efficiency of the cities in the Yangtze River Economic Belt. The Dagum Gini coefficient, Kernel density estimation, and Markov chain analysis were used to examine its spatial pattern and temporal trend, and the key influencing factors were tested using quantile regression. The study found that: (1) The level of ecological efficiency in downstream cities was significantly higher than that in the upper and middle reaches. (2) The overall difference in urban ecological efficiency was large but gradually decreasing. The intraregional difference in urban ecological efficiency was always the largest in the downstream areas, and the largest regional difference was between the upstream and the downstream regions. With the improvement of the level of ecological efficiency, the impact coefficients of factors such as technological innovation and public awareness of environmental protection had gradually decreased, which is an important reason for the reduction of the overall and intraregional differences. The influencing factors in different regions and their roles and directions were different, which provides a certain explanation for the regional differences in urban ecological efficiency. (3) Factors such as technological innovation and economic development level had relatively greater impact on cities of low-level and high-level urban ecological efficiency, which resulted in the trend of polarization and multi-polarization in the downstream and upstream regions, respectively. Factors such as economic development level, resource endowment, and industrial structure had low positive impact coefficients on cities of low, medium, and high levels of urban ecological efficiency, and the trend of urban ecological efficiency shift was not obvious, resulting in the overall and regional “club convergence” and “Matthew effect.” This study used the MinDS model to measure the urban ecological efficiency level, effectively avoided the defects of the traditional DEA model, further decomposed the source of the difference on the basis of its overall spatial difference identification, and predicted its long-term transfer trend. This deepened the understanding of the pattern of spatial and temporal change of ecological efficiency in the Yangtze River Economic Belt, which provides some reference for the coordinated promotion of green development of the Yangtze River Economic Belt.