Bilateral stereotactic lesioning of the nucleus accumbens (NAc) core reduces relapse rates in alcohol-dependent patients but may cause irreversible cognitive deficits. Deep brain stimulation has similar effects but requires costly implanted hardware and regular surgical maintenance. Therefore, there is considerable interest in refining these approaches to develop reversible, minimally invasive treatments for alcohol dependence. Toward this end, we evaluated the feasibility of a reverse pharmacogenetic approach in a preclinical mouse model. We first assessed the predictive validity of a limited access ethanol consumption paradigm by confirming that electrolytic lesions of the NAc core decreased ethanol consumption, recapitulating the effects of similar lesions in humans. We then used this paradigm to test the effect of modulating activity in the NAc using the Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) hM3Dq and hM4Di. We found that increasing activity with hM3Dq had no effect, but suppressing activity with hM4Di reduced alcohol consumption to a similar extent as lesioning without affecting consumption of water or sucrose. These results may represent early steps toward a novel neurosurgical treatment modality for alcohol dependence that is reversible and externally titratable, yet highly targetable and less invasive than current approaches such as lesioning or deep brain stimulation.