An efficient gait recognition approach based on wavelet descriptors (WDS) is proposed. First, gait silhouette is described by WDS. Second, the PCA is applied to reduce the dimensionality of the input feature space. Then, the similarity of training samples and test samples is calculated. Finally, sequences matching are used to get the correct identification rate. By utilizing the proposed approach, the experiments made on CMU database have achieved comparatively high correct identification rate. And related experiment results show that, in the same conditions, the correct identification achieved by WDS is generally higher than the method of Centroid and FDS.