Although information retrieval systems based on the words and phrases in documents offer excellent performance, such systems are limited by their lack of knowledge about documents. When the goal is the management of knowledge to support reuse, information systems must be context aware. The information need that leads to a search for documents has a context, as do the documents of the search space. Only by matching the contexts of the documents, as well as the words and structure of the documents, can knowledge be effectively reused. We describe the SINGLESOURCE system, which has been designed to facilitate knowledge reuse within a group of professional researchers at a major consulting firm.