Background: The relationship between gestational age at birth (GA) and alcohol use measures in early adulthood was examined in a large U.K. community-based birth cohort (Avon Longitudinal Study of Parents and Children). Methods: A series of linear and logistic regression models were used to test for main effects of a continuous measure of GA on a range of alcohol use measures, and moderation of these associations by sex. In addition, mediation analyses assessed the extent to which significant associations between GA and alcohol use operated indirectly, through influences of the parental environment and/or childhood measures of emotional and behavioral health (EBH). Results: Earlier GA significantly predicted never drinking by age 18, but was not associated with other measures of alcohol use behavior among young adult drinkers (i.e., Self-Rating of the Effects of Alcohol, Alcohol Use Disorders Identification Test, or DSM-IV-TR Criteria for Alcohol Dependence). The association between earlier GA and never drinking by age 18 was moderated by sex, such that females born early were less likely to have ever had a drink by age 18. In the full sample, childhood measures of EBH were found to mediate the association between earlier GA and never drinking by age 18. This association was not mediated by parenting factors. Conclusions: Earlier GA is associated with never drinking alcohol in early adulthood, in females. Emotional and behavioral difficulties experienced in early childhood may mediate the relationship between earlier GA and never drinking by age 18.