Due to the information technology which is rapidly developing, digital content is becoming increasingly difficult to handle. This includes images that are kept on digital cameras, CCTV, and medical scanners. Areas such as medical and forensic science are using these databases to do critical tasks which include diagnosing diseases or identification of criminal suspects. However, managing and search the similar images from these databases is not an easy task. Content-Based Image Retrieval (CBIR) is one of the techniques used to manage and search similar images from a database. The performance of CBIR depends on the low level (Color, Texture, and Shape) features. In this paper, a new feature vector to represent the image in terms of low-level features and to improve the performance of CBIR is proposed. The proposed CBIR system was developed based on the combined features of Color, Texture, and Shape Features. Color features are extracted by using Color moments, HSV Color Autocorrelogram Feature (CAF), and Color Layout Descriptor (CLD). Texture features are extracted by Gray Level Co-Occurrence Matrix (GLCM) and Segmentation-based Fractal Texture Analysis (SFTA). Shape features are extracted by Hu Moments. The combined features which are made up of 165 Color Descriptor features, 109 GLCM-SFTA texture features, and 8 shape Hu's moment values are extracted to both query and database images. The extracted feature vector of the query image is compared with extracted feature vectors of the database images to obtain similar images.