ABSTRACT. Objective:This study compared baseline metabolite levels in components of the brain reward system among individuals who remained abstinent and those who resumed hazardous alcohol consumption after treatment for alcohol dependence. Method: Fiftyone treatment-seeking alcohol-dependent individuals (abstinent for approximately 7 days [SD = 3]) and 26 light-drinking nonsmoking controls completed 1.5-T proton magnetic resonance spectroscopic imaging, yielding regional concentrations of N-acetylaspartate, cholinecontaining compounds, creatine-containing compounds, and myoinositol. Metabolite levels were obtained in the following component of the brain reward system: dorsolateral prefrontal cortex, anterior cingulate cortex, insula, superior corona radiata, and cerebellar vermis. Alcohol-dependent participants were followed over a 12-month period after baseline study (i.e., at 7 days of abstinence [SD = 3]) and were classifi ed as abstainers (no alcohol consumption; n = 18) and resumers (any alcohol consumption; n = 33) at follow-up. Baseline metabolite levels in abstainers and resumers and light-drinking nonsmoking controls were compared in the above regions of interest. Results: Resumers demonstrated signifi cantly lower baseline N-acetylaspartate concentrations than light-drinking nonsmoking controls and abstainers in all regions of interest. Resumers also exhibited lower creatine-containing-compound concentrations than abstainers in the dorsolateral prefrontal cortex, superior corona radiata, and cerebellar vermis. Abstainers did not differ from light-drinking nonsmoking controls on baseline metabolite concentrations in any region of interest. Conclusions: The signifi cantly decreased N-acetylaspartate and creatine-containing-compound concentrations in resumers suggest compromised neuronal integrity and abnormalities in cellular bioenergetics in major neocortical components and white-matter interconnectivity of the brain reward pathway. The lack of metabolite differences between abstainers and light-drinking nonsmoking controls suggests premorbid factors potentially contributed to the baseline brain metabolite abnormalities observed in resumers. A LCOHOL-USE DISORDERS (AUDs; i.e., alcohol dependence and abuse) are characterized by a chronically relapsing/remitting course over the lifetime (Dawson et al., 2007;Jin et al., 1998;Miller et al., 2001;Zywiak et al., 2006). The resumption of hazardous levels of alcohol consumption after treatment is common (Donovan, 1996;Maisto and Connors, 2006;Miller et al., 2001;Monahan and Finney, 1996), and appears to be mediated by a complex interplay among genetic, neurobiological, neurocognitive, psychological/psychiatric, and sociodemographic factors (Adinoff et al., 2005;Baler and Volkow, 2006;Bottlender and Soyka, 2005;Bradizza et al., 2006;Glenn and Parsons, 1991;Goodman, 2008;Heinz et al., 2003;Jin et al., 1998;Koob, 2003;Moos and Moos, 2006;Parsons et al., 1990;Sher et al., 2005;Weiss and Porrino, 2002;Zywiak et al., 2006). A considerable amount of research has inve...