Urbanization has amplified regional changes in rainfall associated with global warming, making urban flooding more frequent, but the role of urbanization in this process is still unclear. Based on the Google Earth Engine platform and the random forest algorithm, the land use classification results of the Beijing-Tianjin-Hebei metropolitan area were obtained, and the dynamic division of urban and rural meteorological stations was realized. Then, the “Matthew effect” of rainfall in the region is found, i.e., urbanization decreases light rainfall and increases heavy rainfall. The Geographical Detector model is used to identify the main factors influencing the amount and frequency of rainfall, and hypotheses are proposed accordingly. Finally, based on a total of 470 rainfall events at different stages of urbanization, it is found that light rainfall is mainly distributed at the upwind edge of the city, while rainfall of moderate intensity and above is mainly distributed downwind of the inner cities, and that temperature and subsurface resistance play an important role in the rainfall process. By elucidating the impact of urbanization on rainfall from different perspectives, we propose that city administrators are cognizant of the potential for heatwave hazards associated with a reduction in rain days and flooding caused by increased rainstorm.