Since wavelet transform can not fully describe facial curves features, a face recognition algorithm based on curvelet domain and Kernel Principal Component Analysis (KPCA) is proposed. Using multi-scale, multi-directional curvelet transform to extract image features not only has higher approximation accuracy and better sparse expression, but also can effectively express the singularity along the curve. Then, KPCA is used to project curvelet feature coefficient into the more expressive kernel space. Finally, the nearest method is adopted for classification. The results indicate this algorithm has better effect on image dimension reduction and face recognition rate in the JAFFE face database, ORL face database and FERET face database.