Traditional published norms for neuropsychological tests that do not consider demographic effects can lead to spuriously high false positive rates among low-educated elderly individuals. This problem may be compounded when trying to identify dementia in psychogeriatric patients whose cognitive functioning is also compromised by psychiatric illness. This study investigated the clinical utility of low education neuropsychological test norms to discriminate amongst demented and nondemented psychogeriatric inpatients and healthy community elderly with limited education. Results indicated that the Mattis Dementia Rating Scale (MDRS), the Fuld Object Memory Evaluation (FOME), the Mini-Mental State Examination (MMSE), and a Clock drawing task had high discriminability in differentiating the three groups. Application of demographically corrected norms has important implications for diagnosis and treatment planning, especially when neuropsychological status is complicated by psychiatric illness.