This article examines the ability of Beijing Climate Center Climate System Model (BCC_CSM) in demonstrating the prediction accuracy and the leading modes of the summer precipitation over North Asia (NA). A dynamic-statistic combined approach for improving the prediction accuracy and the prediction of the leading modes of the summer precipitation over NA is proposed. Our results show that the BCC_CSM can capture part of the spatial anomaly features of the first two leading modes of NA summer precipitation. Moreover, BCC_CSM regains relationships such that the first and second mode of the empirical orthogonal function (EOF1 and EOF2) of NA summer precipitation, respectively, corresponds to the development of the El Niño and La Niña conditions in the tropical East Pacific. Nevertheless, BCC_CSM exhibits limited prediction skill over most part of NA and presents a deficiency in reproducing the EOF1's and EOF2's spatial pattern over central NA and EOF2's interannual variability. This can be attributed as the possible reasons why the model is unable to capture the correct relationships among the basic climate elements over the central NA, lacks in its ability to reproduce a consistent zonal atmospheric pattern over NA, and has bias in predicting the relevant Sea Surface Temperature (SST) modes over the tropical Pacific and Indian Ocean regions. Based on the proposed dynamic-statistic combined correction approach, compared with the leading modes of BCC_CSM's original prediction, anomaly correlation coefficients of corrected EOF1/EOF2 with the tropical Indian Ocean SST are improved from 0.18/0.36 to 0.51/0.62. Hence, the proposed correction approach suggests that the BCC_CSM's prediction skill for the summer precipitation prediction over NA and its ability to capture the dominant modes could be certainly improved by choosing proper historical analogue information.