It is important that clinicians who rate global change as part of Alzheimer's disease (AD) clinical drug trials agree on a relevant set of behaviors and information to be considered in formulating their rating. Yet, consensus among raters has been difficult to establish, and inter-rater reliability of clinical global impression of change (CGIC) ratings has been low. In preparation for the development of a new CGIC scale to be used in AD clinical trials, the Alzheimer's Disease Cooperative Study-Clinical Global Impression of Change (ADCS-CGIC), we surveyed clinicians at sites comprising the National Institute on Aging-sponsored ADCS participating centers to identify whether or not consensus regarding CGICs exists. Overall, respondents reported that a CGIC should include an assessment of the patient's function and mental status, a care giver interview, and a standardized set of questions, and it should take approximately 20 minutes per interview. Depending on a patient's level of impairment, raters consider different areas of behavior in formulating a CGIC rating. These findings demonstrate the considerable consensus regarding the CGIC rating process, and were integrated into the design of the ADCS-CGIC, currently in use.