A barrier in outcomes and effectiveness research is the ability to characterize the interventions under review. This has been the case especially in rehabilitation in which interventions are commonly multidisciplinary, customized to the patient, and lack standardization in definition and measurement. This commentary describes how investigators and clinicians, working together, in a major multisite stroke rehabilitation outcome study were able to define and characterize diverse stroke rehabilitation interventions in a comprehensive, yet parsimonious, fashion and thus capture what actually transpires in a hospital-based stroke rehabilitation program. We consider the implications of the study's classification system for a more comprehensive taxonomy of rehabilitation interventions and the potential utility of such a taxonomy in operationalizing practice standards, medical record keeping, and rehabilitation research.