Neuropsychology is poised for transformations of its concepts and methods, leveraging advances in neuroimaging, the human genome project, psychometric theory, and information technologies. It is argued that a paradigm shift toward evidence-based science and practice can be enabled by innovations, including (1) formal definition of neuropsychological concepts and tasks in cognitive ontologies; (2) creation of collaborative neuropsychological knowledgebases; and (3) design of Web-based assessment methods that permit free development, large-sample implementation, and dynamic refinement of neuropsychological tests and the constructs these aim to assess. This article considers these opportunities, highlights selected obstacles, and offers suggestions for stepwise progress toward these goals.