Selecting outcome measures that are both psychometrically sound and sensitive to change is a very important aspect of clinical outcome research. A variety of measures have been introduced in recent years to assess behavioral complications in dementia, but few have been adequately tested in clinical trials. This article provides a discussion of factors to consider in selecting measures, including psychometrics, item content, assessment source, and sensitivity to change. A review of behavioral and psychiatric measures for dementia patients is provided, including measures of general behavioral disturbance, and measures specifically developed for agitation and depression. Each measure's psychometric characteristics, prior use with demented patients, and strengths and weaknesses with regard to treatment outcome research is summarized. The importance of linking measures to the investigators' hypotheses is discussed, along with recommendations for evaluating and selecting outcome measures depending on the needs of the specific investigation.