In data communication the main problem is with security and confidentiality of the contents of communication. Hence, the field of Steganography has been attracted considerable attention from last two decades. Recently Steganography has become most important as it maintains the best property to recover the original data while extracting the contents of an image. Reversible data hiding is a steganographic technique in which the image containing the hidden data is losslessly recovered. Here introduced a new technique which uses sparse representation for reversible data hiding in encrypted images. Compressing the signals and then representing them using dictionary atoms is done by using sparse representation. The proposed work focuses on the patch level sparse representation instead of pixel level for hiding secret data in the encrypted images, because the neighboring pixels (patch) are highly correlated to each other. By using this method, large space for hiding secret message can be achieved.