Abstract The Self-Rating of the Effects of Alcohol (SRE) is a widely used and well-established measure of the level of response to alcohol. Although the SRE has been successfully used in studies of alcoholism etiology, including genetics, studies to date have not compared the self-report and interview formats. The objectives of this study are to: (a) test the predictive utility of the subscales of the SRE in relation to alcohol problems; and (b) test the reliability of the SRE in interview versus self-report formats. A sample of college drinkers (n=446) completed the SRE in a self-report format along with the Alcohol Use Disorders Identification Test (AUDIT). A subset of participants (n=34) returned to the laboratory and completed the SRE in a face-to-face interview format. All subscales of the SRE were robust predictors of alcohol problems accounting for as much as 25% of the variance in AUDIT scores. In addition, scores obtained via self-report and interview-based SRE were highly correlated (r=.70 to .80). Results support the predictive utility of the SRE and provide initial evidence that the self-report and interview formats produce reliable results and may be combined and/or used interchangeably.