Abstract: A human face recognition algorithm based on fast wavelet transform (FWT) and Fisher's linear discriminant (FLD) is presented to meet the requirement of quick face recognition at high rate. In the algorithm the Haar wavelet with scales of 1 and 2 is applied separately to the decomposition of typical human face, then the features are extracted from the original image and the images decomposed on the scales of 1 and 2 by FLD. Consequently, the features extracted are recognized by nearest neighbor classifier. A simulation of the algorithm proposed was done on the basis of ORL (Olivetti Research Lab) face database, and the results showed that the algorithm is able to recognize quickly with high recognition rate.