A prediction model of financial distress based on wavelet neural network for listed companies is proposed. The method of selecting financial ratios of companies is analyzed. The self-adaptive learning rate and momentum coefficient are used to avoid the local minimum point in the training process of wavelet neural network. The prediction results show that compared with multiple discriminate analysis, and Logit and Probit models, the prediction with this method is more accurate and type I and type II errors are reduced significantly.