Abstract: To evaluate applicability of 3 reanalysis datasets of precipitation （CMFD, MSWEP and ERA 5）in the Changjiang River Basin, this paper used the quantitative evaluation indices and the classified statistical indices to evaluate the accuracy of the three reanalysis datasets at multiple scales based on the daily precipitation data of 226 meteorological stations in the Changjiang River Basin from 1979 to 2016. The results show that ERA 5 shows an overestimation phenomenon in the Changjiang River Basin as a whole among the three reanalysis datasets. CMFD can better reflect the actual precipitation change whether it is for daily scale, monthly scale or annual scale accuracy evaluation. In terms of the precipitation detection capacities of different magnitudes, the order is CMFD>ERA5>MSWEP. The performances of CMFD and MSWEP in flood season are better than those in non-flood season. Although ERA5 overestimates monthly precipitation, its annual precipitation distribution trend is consistent with that of gauge-based precipitation. In general, the applicability of CMFD in the Changjiang River Basin is higher than that of MSWEP and ERA5. The research results can provide a basis for the selection of reliable precipitation reanalysis products for related research and applications such as precipitation and hydrological simulation in the Changjiang River Basin.