Outpatients from two primary care clinics in a county hospital were screened for alcohol problems under three conditions. The objective was to determine whether patients would report more drinking and alcohol-related problems if self-report information was gathered in a research setting and/or if patients knew that their doctors would not be informed about their self-reports. The data indicate that patients' self-reports of alcohol consumption or problems were not affected by their knowledge that doctors would know their responses, and that interviews conducted in medical settings did not corrupt self-report data.