The present study investigated discrepancy between reports of depressive symptoms of 36 psychogeriatric patients and their family informants. It also examined factors potentially affecting this discrepancy such as selected characteristics of the patients and their informants, the type of measure assessing depression, and the type of depressive symptoms being assessed. The 15-item Geriatric Depression Scale (GDS-15) and the Clinically Useful Depression Outcome Scale (CUDOS) were completed by the patient, and the informant version of both the GDS-15 and CUDOS were completed by their informant. A sizable discrepancy was found between patient and informant reports of depressive symptomatology; informants reported significantly more symptoms than patients themselves. The discrepancy in reports was greater on the GDS-15 than on the CUDOS. Multiple regression analyses revealed that both patient‟s gender and type of setting (inpatient vs. day hospital) significantly influenced the discrepancy. The highest kappa agreement was obtained on items related to feelings of worthlessness and life satisfaction on the GDS-15, and suicidal ideation and intent on the CUDOS. The study‟s strengths and limitations, implications for clinical practice and research, and directions for future research are discussed.