The general linear model(GLM) is a statistical model of hydrological forecast post-processing.It can reduce biases in raw simulations,preserve skills of raw forecasts,and produce reliable hydrologic ensemble forecasts.In this paper,GLM is used to adjust the daily hydrograph simulations,which are taken from the international model parameter estimation experiment database,and the adjusted results are compared with the observations in order to test GLM's capability of reducing biases.Results show that the continuous ranked probability scores of the hydrological ensemble forecasts are all below 0.5,proving that the ensemble forecasts are reliable.The experimental indexes of the adjusted simulations,such as mean,standard deviation,root-mean-square error,are more approximate to the observations than those of the raw simulations.Even to the simulations which contain fewer biases,GLM still can improve them.