This paper proposes a Dictionary-Learning Sparse Representation framework (DLSR) to deal with face pose estimation in noise, bad illumination and low-resolution cases. Sparse and redundant modelling of data assumes an ability to describe signals as linear combinations of a few atoms from a pre-specified dictionary. As such, the choice of the dictionary that sparsifies the signals is crucial for the success of this pose estimation problem. The proposed approach models the appearance of face images from the same pose via a sparse model which learns the dictionary D from a set of image patches with the objective to minimize the reconstruction error of the target image, in order to coincide with the pose classification criterion. Then, the combination of the trained dictionaries of all pose classes are used as an over-complete dictionary for sparse representation and classification. Experimental results demonstrate the effectiveness of the proposed Dictionary-Learning Sparse Representation framework for treating the pose classification in dynamic illumination condition and low-resolution images.