Abstract: Statistical properties of annual and seasonal daily maximum precipitation in terms of nonstationarity, upper tail properties, and impact of tropical cyclones (TCs) were analysed, using data from 728 meteorological stations across China and TCs tracks data covering a period of 1951–2014. The change points and trends detected by stepwise regression method and modified Mann–Kendall, respectively, provide limited evidence of the violation of the stationarity assumption in both annual and seasonal extreme precipitation, and we cannot make conclusive statements about the presence of a possible climate change signal. For stationary stations, the ratio of scale to location parameters of generalized extreme value (GEV) distribution shows exponential for annual and seasonal extreme precipitation in some sub‐regions. On the other hand, annual and seasonal maxima point to an unbounded‐above and heavy‐tail behaviour, and discussion of these upper tail properties in light of mixtures of different rainfall‐generating mechanisms is provided. In addition, anomalous values of the GEV shape parameter are associated with TCs in controlling the upper tail of annual maxima on China's coastal areas with strong spatial heterogeneities.