Background: Women who terminate pregnancies drink more subsequent to the pregnancy than women who give birth, including women who give birth after seeking to terminate a pregnancy., Methods: Data are from the Turnaway Study, a prospective, longitudinal study of 956 women who sought to terminate pregnancies at 30 U.S. facilities. This paper focuses on the 452 women who received terminations just below facility gestational limits and 231 who were denied terminations because they presented just beyond facility gestational limits. This study examined whether baseline characteristics moderate the relationship between termination and subsequent binge drinking and whether stress, feelings about the pregnancy, and number of social roles mediate the relationship., Results: Only having had a previous live birth modified the termination-binge drinking relationship. Among women with previous live births, binge drinking was reduced among women carrying to term compared to terminating the pregnancy. Among women who had not had a previous live birth, however, the reduction in binge drinking among those denied termination was not sustained over time, and binge drinking of those who had and had not had terminations converged by 2.5 years. Neither stress, negative emotions, nor social roles mediated effects on binge drinking. Positive emotions at one week mediated effects on binge drinking at six months, although positive emotions at two years did not mediate effects on binge drinking at 2.5 years., Conclusions: Higher levels of binge drinking among those who terminate pregnancies do not appear due to stress or to negative emotions. Only parous women - and not nulliparous women - denied terminations experienced sustained reductions in binge drinking over time., (Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.)