Until now, no self-rated depression scale had been validated as a screening measure for major depression in the older patient hospitalized with medical illness. The present report establishes the validity of two brief, easily administered depression screening tests, the Geriatric Depression Scale (GDS) and the Brief Carroll Depression Rating Scale (BCDRS), in this population. Structured psychiatric interviews were performed and self-rated depression measures administered to 128 men, aged 70 and over, consecutively admitted to medical and neurological services of a VA hospital. The GDS and BCDRS were both shown to have high sensitivity and specificity for detecting major depression in this setting. Optimal cut-off scores determined by the receiver operating curve characteristics of these tests were 11 for the GDS and 6 for the BCDRS. At a cutoff score of 11, the GDS had a sensitivity of 92%, a specificity of 89%, and a negative predictive value of 99%; lowering the break point to 8 did not increase sensitivity. At a cutoff score of 6, the BCDRS achieved a 100% sensitivity, 93% specificity, and 100% negative predictive value. Whether clinicians decide to implement either of these depression screens in their practice will depend to a large degree on the importance ascribed to the detection of these disorders and on attitudes toward the benefits of treatment.