The Clinician Interview-Based Impression of Change, plus carer interview (CIBIC-Plus), is widely used in antidementia drug trials. It comprises Likert scales for disease severity and changes, and written accounts summarizing semistructured interviews evaluating behavior, cognition, and function. Studies using the CIBIC-Plus have focused on the numeric scores to the exclusion of the textual data. Our study explored both sets of data to evaluate whether the CIBIC-Plus written data supported (a) the clinicians' global evaluation of patients' changes during treatment, and (b) the emergence of consistent treatment effects. The global (numeric) scales of change were inconsistently supported by the textual data provided in the CIBIC-Plus. No consistent treatment effects were noted. Methodological problems presently limit the retrospective use of the CIBIC-Plus textual data. Improved standardization of note-taking in the CIBIC-Plus textual data may allow for a better understanding of the typical profiles and clinical importance of changes seen in the course of dementia treatment.