Abstract: Background Clinical measures continue to be used as primary endpoints for disease-modifying trials for Alzheimer‘s disease (AD). Currently, two co-primary endpoints must be specified, which measure cognitive and functional impairments. Generally, the Alzheimer‘s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog) is one of the co-primary endpoints, but high variability in this measure results in large sample sizes. We evaluated the psychometric properties of the Clinical Dementia Rating-Sum of Boxes (CDR-SB) to assess its suitability as a single primary endpoint as an alternative to the traditional co-primary approach. Methods Internal consistency, structural and convergent validity, and 2-year internal and external responsiveness of the CDR-SB were assessed in 667 very mild to moderate (global Clinical Dementia Rating, 0.5–2) AD patients from the REAL.FR (Reseau sur la Maladie d'Alzheimer Francais) study. Results The CDR-SB showed good internal consistency (Cronbach‘s alpha = 0.88), and acceptable structural (separate ”cognitive“ and ”functional“ factors) and convergent validity. Variability in mean changes over time was low, leading to excellent internal responsiveness (effect size = 1.2; standardized response mean = 1.17 at 2 years) and smaller sample sizes as compared with the ADAS-Cog. External responsiveness was acceptable when compared with ”clinically meaningful“ changes on the Activities of Daily Living scale but only borderline acceptable when compared with the ADAS-Cog and Instrumental Activities of Daily Living. Levels of missing data and floor/ceiling effects were low. Conclusions The CDR-SB measures cognitive and functional impairment simultaneously, and has excellent 2-year internal responsiveness. This makes it a promising candidate as a sole primary endpoint for AD trials, although more work is required to determine the clinical relevance of CDR-SB changes, and its usefulness as an endpoint at other disease stages.