Recently, steganography and steganalysis have been received an increasing attention due the nature of our modern societies which depends on exchanging information on a large scale. Steganography is the art of communication through sharing secret messages by embedding them into useless cover messages. The cover message can be an image, audio, or video file. On the other side, the steganalysis techniques are concerned with discovering the existence of steganography. This paper presents a specific image steganalysis technique with main objective is to detect the existence of steganography made by the least significant bit (LSB) technique in a certain image. The proposed approach extracts the gray level co-occurrence matrix (GLCM) as salient features which capable to distinguish a stego image from a non-stego one using a Back-Propagation (BP) classifier at the classification phase. Experimental results on standard datasets that consists of 297 images are encouraging. The proposed method is robust and high accuracy level has been achieved.