Objective: This study sought to assess the body mass index (BMI)-alcohol relationship among a US national sample of college students. Design: Secondary data analysis using the Fall 2011 National College Health Assessment (NCHA). Setting: A total of 44 US higher education institutions. Methods: Participants included a national sample of college students (N?=?27,774). BMI, frequency of alcohol consumption and binge drinking were measured based on responses to items in the NCHA. T-tests and two-way analysis of variance (ANOVA) identified differences in BMI based on drinking status and sex. Results: Drinkers exhibited significantly lower BMIs than non-drinkers, t(26,968)?=?5.55, p?<?0.001. Binge drinkers exhibited significantly lower BMIs than non-binge-drinkers, t(27,094)?=?2.64, p?<?0.001. A significant interaction for drinking frequency and sex existed, F(5, 20,434)?=?4.121, p?=?0.001, but no interaction effect was present for sex and binge drinking frequency. Respondents drinking 1-20 days/month had lower BMIs compared to non-drinkers. Furthermore, frequency of past month drinking (ß?=?-0.08, p?<?0.001) made a unique contribution to explaining BMI, such that BMI decreased as the days one consumed alcohol increased. Conclusion: Non-drinkers exhibited significantly higher BMIs than drinkers, regardless of whether consumption was measured as drinking status, frequency of past 30 day use, or binge drinking frequency. Findings build on previous literature which has produced mixed results depending on the operationalisation of alcohol use. Future research into "why" differences exist in BMI among drinkers versus non-drinkers is warranted.