Abstract: This paper proposes a precipitation path identification method that accounts for factors such as precipitation cloud speed, the spacing between meteorological stations, and the overall direction of precipitation movement. The method consists of three steps: spatial–temporal lag correlation detection of precipitation at neighboring stations, principal component analysis, and curve fitting. It was applied to identify and analyze summer precipitation paths in Beijing using data from automatic meteorological stations collected between June and August from 2010 to 2020. The results indicate that the shortest distance between adjacent stations in Beijing is 0.1–7.7 km, with an average of 3.3 km. A total of 18 primary summer precipitation paths, generally oriented from southwest to northeast. The path lengths range from 10 to 125 km (average 33 km), and the durations along the paths vary from 13 to 148 min (average 43 min). Precipitation paths passing through urban areas tend to be longer, whereas those in mountainous regions are shorter and mostly located on windward slopes. Notably, precipitation paths do not necessarily traverse zones of heavy precipitation. Precipitation amounts often increase or decrease progressively along the paths. Overall, the precipitation path identification method is reliable and practical. Its results can support the spatial deployment of artificial weather modification operations and provide valuable reference for short-term and nowcasting applications.