Diagonal maximum scatter difference (DiaMSD) method for face recognition is proposed in this paper. This novel algorithm is developed based on two techniques, i.e., maximum scatter difference (MSD) and diagonal face images based projection. The DiaMSD method is not only computationally more efficient but also more accurate than the one dimensional (vector-based) MSD method in extracting the facial features for human face recognition. Extensive experiments are performed to test and evaluate the new algorithm using a subset of the FERET face databases. Experimental results show the effectiveness of the proposed method (DiaMSD).