The research of face classification achieves people's attention with increasing of face database. In front-view images, chin contour is a relatively stable shape feature, but it is hard to extract by conventional edge-detection methods due to the complexity of grayscale distribution in chin area. A novel algorithm for chin contour extraction and a face classification method using chin contour are presented. Three chin templates were constructed and the template matching was done. Chin classification was achieved according to the matching results, and then face classification results were got. Based on prior distribution and local decision, the possible points of chin contour were got using the algorithm. The false points were deleted by filtering and the other points were joined to get chin contour. Then we classified them by template matching. Experimental results show that we can get good classification results of chins and faces by this algorithm, which is good enough to improve the recognition rate and speed on huge face database.