This author identifies a critical problem in the wholesale adoption of patient classification systems (PCSs) in nursing departments. She suggests that without attention to their reliability and validity for the specific institution's standards and practice, PCSs may frustrate, rather than assist, the nursing administrator's efforts to predict staffing needs. Here she describes an “ideal” PCS and compares it with the reality of today's systems. She then suggests ways that nursing administrators can move closer to this ideal system and make PCSs more useful management tools.