OBJECTIVE Use of cognitive, psychiatric and behavioral domains to assess the effectiveness of a pharmacological or behavioral intervention in the treatment of neurodegenerative disorders (e.g., Alzheimer disease) aids the identification of specific types of impairment or distress in behavioral status and quality-of-life issues in this population. With confirmatory approaches to subscale development readily available, we can obtain a more precise understanding of the sub-components of a scale, potentially providing the basis for selecting behavioral and/or quality-of-life outcome measures that may be more sensitive to the effects of pharmacological or behavioral interventions. METHODS The authors illustrate the use of a confirmatory factor-analytic approach to verify scale sub-domains of the Neurobehavioral Rating Scale (NBRS) in elderly patients with dementia. With data collected from two groups of patients being treated for significant psychiatric and behavioral symptoms, authors investigated the relationships among scale items in order to test several competing models, including a general one-factor model, a first-order multifactor model, and a second-order factor model. RESULTS The first-order model, with seven factors, provided the best fit to the correlations among items in the NBRS, indicating the multidimensionality of problematic behaviors and symptoms exhibited by dementia patients. CONCLUSION Authors advocate the use of a confirmatory factor-analytic approach to verify scale sub-domains in other, more widely used rating scales for patients with dementia.