Abstract: Reservoir overtopping occurs under the conditions of the sudden flood is a common situation for dam failure, and how to analyze the reservoir overtopping risks seems so important in the overtopping event prediction. On the basis of the analysis of the characteristics of the reservoir overtopping risks under the conditions of the sudden flood, we define an overtopping limit state function to establish a maximum entropy model for analysis of the sudden flood overtopping limit state risks, and to determine the probability density function of flood waves overtopping risks. The method of BP neural network has been adopted to fit the reservoir overtopping limit state function, calculate the probability density function of the reservoir waves overtopping risks by using a quadratic fourth-order moment method, give the solving steps of the probability of the reservoir waves overtopping risks, and further calculate the probabilities of the reservoir waves overtopping risks, thus providing a scientific basis for the prediction and prevention of the reservoir overtopping. Finally, by calculating and analyzing a small and medium mountain reservoir, the analysis results show that the calculated results of the maximum entropy method are very close to those of the other methods, and that the model has a higher computational efficiency.