Although alcoholism and depression are highly comorbid, treatment options that take this into account are lacking, and mouse models of alcohol (ethanol (EtOH)) intake-induced depressive-like behavior have not been well established. Recent studies utilizing contingent EtOH administration through prolonged two-bottle choice access have demonstrated depression-like behavior following EtOH abstinence in singly housed female C57BL/6J mice. In the present study, we found that depression-like behavior in the forced swim test (FST) is revealed only after a protracted (2 weeks), but not acute (24 h), abstinence period. No effect on anxiety-like behavior in the EPM was observed. Further, we found that, once established, the affective disturbance is long-lasting, as we observed significantly enhanced latencies to approach food even 35 days after ethanol withdrawal in the novelty-suppressed feeding test (NSFT). We were able to reverse affective disturbances measured in the NSFT following EtOH abstinence utilizing the N-methyl D-aspartate receptor (NMDAR) antagonist and antidepressant ketamine but not memantine, another NMDAR antagonist. Pretreatment with the monoacylglycerol (MAG) lipase inhibitor JZL-184 also reduced affective disturbances in the NSFT in ethanol withdrawn mice, and this effect was prevented by co-administration of the CB1 inverse agonist rimonabant. Endocannabinoid levels were decreased within the BLA during abstinence compared with during drinking. Finally, we demonstrate that the depressive behaviors observed do not require a sucrose fade and that this drinking paradigm may favor the development of habit-like EtOH consumption. These data could set the stage for developing novel treatment approaches for alcohol-withdrawal-induced mood and anxiety disorders.