the content based color image retrieval has great interest nowadays. In this paper, we have proposed a new approach for such systems. The proposed approach employs two main phases in train and test. At the first phase, color features are extracted based on RGB color space and texture are represented by Texton. In the second phase, the images are clustered based on extracted features using a Self-Organizing Map (SOM) neural network, the experimental results showed the method performance versus related works.