The randomly weighted prior bayesian method was investigated to estimate the parameters of P-III distribution. The parameters prior distribution is determined by randomly weighted method, employing the AM algorithm to sampling analysis for the poste riori distribution of parameters, and compared with the traditional hydrological frequency analysis methods, such as Moments method, Maximum Likelihood method and Probability Weighted Moment method. The examples show that the parameters obtained by the Bayesian AM algorithm has the minimum sum of squares of deviations between the measured samples and design values under their frequencies. The AM algorithm is a feasible hydrological frequency analysis approach.