This paper presents a new interface for searching and browsing large image collections, which allows users to navigate along conceptual dimensions that describe the images. The interface uses hierarchical faceted metadata and dynamically generated query previews to seamlessly integrate category browsing with keyword searching. A usability study with 32 art history students explored a collection of 35,000 fine arts images and compared this approach to a standard image search interface. The study found that 90% of participants preferred the metadata approach, 97% said it helped them learn more about the collection, 75% found it more flexible, and 72% found it easier to use than a standard baseline system. These results indicate that a category-based approach is a successful way to provide access to image collections. The paper discusses related work, the faceted metadata, the category-based interface design, the baseline interface, and the study design and results. It concludes with a discussion of the larger lessons that can be drawn from this effort. The study found that users preferred the faceted category interface over the baseline interface, suggesting this is a promising direction for image search interfaces. The interface was designed to support search usability guidelines while avoiding negative consequences like empty result sets or feelings of being lost. The interface allows users to navigate along conceptual dimensions, and results can be selected by keyword search, by pre-assigned metadata terms, or by a combination of both. The interface was tested with 32 participants, and the results showed that the faceted category interface was preferred over the baseline interface. The study also found that users were more likely to use the faceted category interface in the future and felt more familiar with the collection. The interface was found to be more useful and flexible than the baseline interface, and users were more satisfied with the results. The study also found that the faceted category interface was more effective for complex searches, and users were more likely to use it for multiple-facet queries. The study concluded that a category-based approach is a successful way to provide access to image collections.This paper presents a new interface for searching and browsing large image collections, which allows users to navigate along conceptual dimensions that describe the images. The interface uses hierarchical faceted metadata and dynamically generated query previews to seamlessly integrate category browsing with keyword searching. A usability study with 32 art history students explored a collection of 35,000 fine arts images and compared this approach to a standard image search interface. The study found that 90% of participants preferred the metadata approach, 97% said it helped them learn more about the collection, 75% found it more flexible, and 72% found it easier to use than a standard baseline system. These results indicate that a category-based approach is a successful way to provide access to image collections. The paper discusses related work, the faceted metadata, the category-based interface design, the baseline interface, and the study design and results. It concludes with a discussion of the larger lessons that can be drawn from this effort. The study found that users preferred the faceted category interface over the baseline interface, suggesting this is a promising direction for image search interfaces. The interface was designed to support search usability guidelines while avoiding negative consequences like empty result sets or feelings of being lost. The interface allows users to navigate along conceptual dimensions, and results can be selected by keyword search, by pre-assigned metadata terms, or by a combination of both. The interface was tested with 32 participants, and the results showed that the faceted category interface was preferred over the baseline interface. The study also found that users were more likely to use the faceted category interface in the future and felt more familiar with the collection. The interface was found to be more useful and flexible than the baseline interface, and users were more satisfied with the results. The study also found that the faceted category interface was more effective for complex searches, and users were more likely to use it for multiple-facet queries. The study concluded that a category-based approach is a successful way to provide access to image collections.