Difficulties in detecting depressive illness in the elderly have recently led to the development of screening tests which focus on depressive symptoms typical of older persons. The present study examined the utility of one such test, the Short GDS, in a heterogeneous elderly population. The sample consisted of 20 clinically depressed elderly, matched with 20 elderly from a Jerusalem neighborhood sample. The test was found to significantly discriminate depressed and normal community-dwelling elderly. Sensitivity and specificity were found to be inadequate for large-scale screening. This was hypothesized to be due to the use, in the present study, of depressed persons on antidepressant medication, cultural variations in the interpretation of items, the treatment of missing data, and the variance in community sample. The Short GDS has utility for use in a two-stage screening process.