Although the introduction of explicit diagnostic criteria and rule-based classifications, such as Research Diagnostic Criteria, Diagnostic and Statistical Manual Ed 3 and Ed 4, has dramatically influenced teaching and research psychiatric practice, it has significant limitations. As new knowledge is developed with genetic, imaging and metabolomic technologies, a method to incorporate this research in a systematic manner with current classification systems is needed. The current approach, which is essentially nominalist in character, has to be developed where new data and new concepts of disease can be integrated and tested. Examples of how this could happen is shown in the context of Alzheimer's disease and subcortical ischemic disease. It is likely that a standardized approach that can develop and modify classification systems in a timely manner, based on science and free of societal and political influence, can enhance research, teaching and clinical practice.