<p>This paper takes the Yangtze River Delta as the research object. Firstly, based on the exploratory spatio-temporal data analysis (ESTDA) framework, the dynamic evolution of water pollution emission intensities is characterized, and then, through the convergence model, it identifies the convergence of the emission intensities and its driving factors. The following results are obtained: (1) During the study period, the emission intensities of chemical oxygen demand (COD) and ammonia nitrogen (NH₃–H) decrease significantly, and their spatial distribution shows a concave pattern of “low center and high edge”, with spatial concentration and regional imbalance highlighted, and the spatial paths coexisted with locking and transition. (2) The water pollution emission intensities present prominent spatial autocorrelation characteristics, and the local spatial structure has strong stability. Both water pollutants have synergistic integration and strong spatio-temporal cohesion, and there is transition inertia. (3) There is no σ-convergence, but there are significant absolute β-convergence, and conditional β-convergence of water pollution emission intensities in the Yangtze River Delta, with the conditional convergence faster than the absolute convergence, and the convergence of NH₃–H faster than COD. The convergence of both of them is significantly affected by the intensity of green technology and foreign investment, and the emission intensity of NH₃–H is affected by urbanization and the regulation of the government at the same time. The spatial spillover of neighboring cities significantly affects the local emission intensities.</p>