Patients with medical or social problems related to alcohol abuse or alcohol dependence frequently are seen in the health care delivery system. Direct and indirect screening methods are available, but contemporary data describing their combined effectiveness have not been reported. We determined the sensitivity of (a) an empirically derived, item-weighted, relatively "indirect" screening measure for alcohol misuse (the Common, Alcohol, Logistic [CAL] scale for the MMPI), (b) a "direct" approach to screening for alcohol misuse (the Self-Administered Alcohol Screening Test [SAAST]), and (c) their combined application. We obtained the responses to both measures from an archival sample of 361 medical patients (89 women, 272 men) who had been diagnosed with alcohol dependence by DSM-III-R criteria. The CAL scale and the SAAST both demonstrated high sensitivity. However, results favored the simultaneous use of a direct-in this study, the SAAST-and an indirect-in this study, the CAL scale-screening procedure.