In this paper, we present a fully automated approach for ear recognition based upon sparse representation. In sparse representation, features extracted from the training data of each subject are used to develop a dictionary. In this work, Gabor filters are used for feature extraction. Classification is performed by extracting features from the test data and using the dictionary for representing the test data. The class of the test data is then determined based upon the involvement of the dictionary entries in its representation. Experimental results conducted on the University of Notre Dame (UND) collection G dataset, containing large appearance, pose, and lighting variability, yielded a rank-one recognition rate of 98.46%. The proposed system outperforms the method described in , which achieves a recognition rate of 96.88% when evaluated on the same dataset. Moreover, the proposed system was evaluated on a greater number of test images per subject, demonstrating its robustness.