Abstract: Eating disorders (EDs) are among the deadliest psychiatric disorders but are underdetected in health care settings, and the majority of people with these conditions do not get treatment for them. There is a need for well-validated and brief screening measures of EDs to aid in early detection and intervention. We compared the performance of two existing brief screeners in a sample of U.S. military veteran men and women, a population in which few evaluations of ED measures have been conducted. We investigated performance of these measures in the full sample and separately in the male and female subsamples. Participants completed a survey assessing EDs and related constructs. A subsample of potential cases and controls completed a diagnostic interview (n = 178). We conducted receiver operator curve analyses to determine cut scores on the two ED screening measures that best discriminated between cases and controls. We identified cut scores on the Eating Disorder Examination-Questionnaire 7 (EDE-Q7) and the SCOFF (the acronym corresponds to the questions) in the full sample and separately among men and women. Model accuracy was fair to considerable for both measures, although internal consistency was low for the SCOFF. The EDE-Q7 performed fairly well among both men and women in this veteran sample. Brief and effective screening makes it more likely that providers will identify and potentially treat EDs in vulnerable populations.