In order to explore the applicability potential for China’s reanalyzed meteorological dataset CN05.1 in watershed hydrological modeling, the SWAT hydrological model was driven by CN05.1 dataset and traditional meteorological station data, respectively, by taking the Kaidu River basin the as research area in this paper. The comparison between the two simulation results was carried out by using evaluation factors such as coefficient of determination (*R* ^{2}), Nash efficiency coefficient (NSE) and relative error (*R*_{e}). Finally, two precipitation data correction methods were used to correct the CN05.1 precipitation data and the hydrological simulation results were evaluated. The results are as follows. (1) CN05.1 meteorological data has certain applicability in the hydrometeorological simulation of the Kaidu River basin. (2) Hydrological simulation based on SWAT model shows that CN05.1 data-driven hydrological simulation accuracy is higher than traditional meteorological station data, and the *R* ^{2} over the period of calibration (1995-2005) and verification period (2006-2016) is 0.81 and 0.73, NSE is 0.81 and 0.72, and *R*_{e} is -0.97% and 0.39%, respectively. (3) The two kinds of correction data can both simulate the process of runoff change well, but runoff simulation based on spatial relationship correction method shows better effect, and the *R* ^{2} and NSE are above 0.72, |*R*_{e}| is less than 1.7%. Therefore, the corrected CN05.1 precipitation data can compensate for the lack of data in the traditional meteorological station and the unsatisfactory peak value of the CN05.1 meteorological data in the runoff simulation to a certain extent.