Objective: The performance on standard clinical and psychometric assessments of eight elderly individuals with major unipolar depression alone and seven with depression plus mild senile dementia of the Alzheimer type was compared with that of 41 nondepressed subjects suffering from very mild senile dementia of the Alzheimer type, 66 with mild senile dementia of the Alzheimer type, and 83 age-matched subjects without senile dementia. Method: Subjects with depression alone, depression plus mild senile dementia of the Alzheimer type, and very mild and mild senile dementia of the Alzheimer type met strict inclusionary and exclusionary criteria. A 90-minute semistructured interview, including several brief standardized clinical scales, was used to assign a Clinical Dementia Rating to each subject according to published guidelines, and each subject was given a 2-hour psychometric test battery. Data were analyzed by one-way multivariate analysis of variance to ascertain if there was an effect of group on clinical and psychometric test scores. Results: The eight depressed subjects without concurrent dementia performed as well as the 83 nondepressed subjects without dementia on most clinical measures; however, their performance on most psychometric measures closely resembled that of the 41 nondepressed subjects with very mild dementia. The performance of the seven subjects with depression plus mild dementia was comparable to that of the 66 nondepressed subjects with mild dementia on most clinical and psychometric measures. Conclusions: Although depressed subjects performed as well as subjects without dementia on many clinical assessments, psychometric testing was not able to distinguish depressed subjects from those with very mild senile dementia of the Alzheimer type. This demonstrates the need for careful psychiatric evaluation before interpreting deficits on psychometric tests as indicating the presence of very mild senile dementia of the Alzheimer type.