This study examined the accuracy of the Diagnostic Interview Schedule (DIS) for indicating alcohol use disorder in a sample of patients hospitalized for depression. The Michigan Alcoholism Screening Test (MAST), with its established validity, was considered a good criterion against which to evaluate the DIS, and preferable to the clinician-assigned diagnosis in this regard. The rates of alcoholism in the sample were 31, 33, and 22.5% as yielded by the DIS, MAST, and physician's diagnosis of alcohol disorder, respectively. (The lower rate for physician's diagnosis may be due to the physician's not applying this diagnosis to recovered or currently abstinent alcohol patients.) Using the MAST's standard cutoff of five points as indicative of alcoholism, agreement with the DIS occurred in 91% of the cases, corresponding to a product moment coefficient of. 79. It was concluded that the DIS alcoholism scale could be used, with reasonable confidence in its validity, for assessing alcoholism in psychiatric settings.