A screening instrument based on the DSM-III-R was developed in order to differentiate aged persons with or without a DSM-III-R mental disorder. The usefulness of this instrument was investigated in six geriatric institutions with 171 residents aged 65 to 93 years in Finland. The nursing staff independently rated the mental status of residents; researchers (a psychiatrist and general practitioner) made a diagnosis based on DSM-III-R criteria. The majority of the residents (mean 69%, varying from 33% to 84% in different institutions) had a DSM-III-R mental disorder. This screening instrument functioned well in differentiating residents without a mental disorder from those with DSM-III-R mental disorder: The nursing staff's agreement was 90%, kappa.71, sensitivity .90, and specificity .88. However, only half the cases of depression were diagnosed correctly: Sensitivity was .50 and specificity .95. With a more specific 9-class diagnostic instrument based mainly on DSM-III-R categories, agreement was only moderate (68%) and kappa .59. The instrument developed in this study seemed to be a useful screening method when applied by the staff, but accurate diagnostics require psychiatric knowledge, especially of depression.