Abstract: To study the probability hydrological forecast and its application value,the Bayesian statistic forecast theory is adopted to formulate the reservoir middle-and long-term runoff forecast model,which describes the uncertainty of hydrological forecast by the distribution function.The gray correlation prediction model for the meteorological factors is presented to count the uncertainty of the input factor.The real-time meteorological information and history hydrological data are coupled effectively,which breakthrough the limitations on the information utilization and samples concerning the determined forecast method and improve the accuracy of hydrological forecast.The established model is tested on the runoff forecast of the Feng-man reservoir.The application results fully show that the superiority is apparent compared with the conventional forecast method.Not only does the model benefits the user to consider the uncertainty quantitatively in decision-making,but ulso improves the forecast accuracy expectatively.