Rainfall erosivity index (EI30) is an important factor used to characterize the effect of rainfall on soil erosion based on storm intensity. Due to the scarcity of breakpoint rainfall data, many scholars had used different fixed time intervals to deduce EI30 in different areas of the world. However, whether there were differences in rainfall erosivity calculated from rainfall data at various time intervals, how the differences were converted, and whether the functional relationship between rainfall erosivity and sediment erosion changes need to be further studied. In this study, 258 storm events collected from a soil and water conservation experimental station in the humid area of southern China were used in the calculation of maximum 30-min rainfall intensity (I30), total kinetic energy of storm (E), and EI30 at different fixed time intervals of 1-, 5-, 10-, 15-, 30-, and 60-min, respectively. At the same time, 86 rainfall erosion sediment yield data of the station were obtained and used them to calculate the relationships between rainfall erosivity and erosion sediment at each time interval. The results showed that the longer the time interval, the greater the underestimation value of I30, E, and EI30. Compared with E value, I30 value played a dominant role in underestimating EI30 value using coarser fixed-interval rainfall data. Nevertheless, there is a significant proportional relationship for EI30 with each other at different fixed time intervals, as well as I30 and E. The basic relationships between EI30 and soil erosion did not change with time interval, but the slope coefficient varied and needed to be calibrated. In order to accurately estimate the rainfall erosivity, it is recommended that the fixed time interval of the rainfall data should not be greater than 30 min.