1. Objective(s): Despite having lower rates of drinking and heavy drinking than European Americans (EAs), African Americans (AA) have significantly higher rates of mortality from a variety of alcohol-related conditions, including liver cirrhosis, accidents, and violence. The current proposal aims to improve alcohol treatment in AA Veterans, who comprise 12% of the Veteran population. 2. Research Design: The proposed study is a two-arm, randomized 12-week, parallel-groups comparison of topiramate versus placebo to reduce the frequency of heavy drinking days and increase the number of abstinent days in 160 AA patients with AUD. 3. Methodology: The following specific aim is used to direct the methods: Specific Aim 1. To test the efficacy of topiramate (TOP) 200 mg/day in reducing the frequency of heavy drinking and increasing abstinent days in African- American (AA) patients with alcohol use disorder (AUD). We hypothesize that, as in European-Americans (EAs), AA subjects receiving TOP will report fewer heavy drinking days (HDDs) and more abstinent days than those receiving placebo (PLA). The analyses make use of all data provided by all patients to estimate models to test the TOP effect on days of heavy drinking and abstinent days. Power for the contrasts will be determined by the patterns of outcomes in the final six weeks, so power estimates are based on weeks 7 through 12 as a 6-week trial, with adjustments for loss to attrition between baseline and week 6. Based on our prior study (Kranzler 2014), we anticipate 92% retention through the first six weeks for each group, yielding 74 available per group at the end of week 6, an additional 4% loss due to dropout across the final six weeks, and a within-subject correlation of about 0.6. The methods of Hedeker et al (1999) show that we will have 80% power for a TOP main effect size of d=0.40, 0.43, and 0.46, for within- subject correlations of 0.5, 0.6, and 0.7, respectively, at a corrected alpha level of 0.025. 4. Impact/Significance: The proposal is innovative in that it will focus on AAs with AUD, an understudied and underserved population for whom no such data currently exist. Given the far-reaching effects of AUD and its high prevalence among veterans, added evidence based treatments may realize reduced health care costs from unnecessary ED visits and reduced complications of illnesses such as hepatitis C and congestive heart failure.