In this paper, we attempt to address the inadequacies of query-based image retrieval, and argue the importance of structured browsing and navigation methods. We propose to build a hierarchical classification model, which is the basic structure for image browsing and navigation. Currently a prototype system has been implemented, called CHROMA. CHROMA provides both a visual tool to browse or navigate through the image collection, which can be used in combination with existing query-based methods. In addition, by measuring the required iterations of thumbnail browsing we evaluate the improvements in retrieval effectiveness of the browsing and navigation approach.