The Columbia University Scale for Psychopathology in Alzheimer's disease is a new screening instrument developed for use by clinicians or trained lay interviewers. Interrater reliability was established between a psychiatrist and a lay interviewer in 20 patients. In an independent sample of 91 outpatients with very mild to moderate probable Alzheimer's disease, caregiver informants reported that depressed mood was common (46.2%) but rarely persistent (2.2%), and that sleep disturbance occurred frequently (41.8%) but was never severe (0%). There were significant but weak associations between the presence of specific subtypes of delusions and severity of dementia. Although a variety of delusional symptoms were reported, they were frequently transient and patients often accepted the truth if corrected by the caregiver. As a result, few patients met broad or narrow operational criteria used to define delusions. Prior studies may have overestimated the prevalence of psychotic features in Alzheimer's disease by not employing standard definitional criteria. The findings also indicate that new methodology such as that employed in this instrument needs to be evaluated more widely.