Intra-storm temporal distributions of precipitation are important for infiltration, runoff, and erosion process understanding and models. A convenient and established method for characterizing precipitation hyetographs is the use of non-dimensional Huff curves. In this study, 11,801 erosive rainfall events with 1 min resolution data collected over 30 to 40 years from 18 weather stations located across the central and eastern parts of China were analyzed to produce Huff curves. Each event was classified according to the quartile period within the event that contained the greatest fraction of rainfall. The results showed that 38.3% of events had the maximum rainfall amounts in the first quartile, followed by the second (26.8%), third (22.4%), and fourth (12.5%) quartiles. Quartile I and II events were generally characteristic of shorter duration and heavier intensity events. Quartile I events averaged 23% shorter durations than quartile IV events, whereas the mean intensity (Iavg), mean maximum 30 min intensity (I30), and mean rainfall erosivity index (EI30) were 1.71, 1.22, and 1.23 times greater, respectively, than those for quartile IV and were significant at a 5% level based on two-sample t-tests. The proportion of quartile I events was less for events of longer duration, whereas the proportions of quartile III and IV events were greater. Two-sample Kolmogorov-Smirnov tests suggested that regional Huff curves can be derived for the central and eastern parts of China. Regional Huff curves developed in this study exhibited dissimilarities in terms of the percentages of storms for different quartiles and the shapes of the curves compared to those reported for Illinois, peninsular Malaysia, and Santa Catarina in Brazil.