ABSTRACT. Objective:Babor's A/B typology characterizes alcoholdependence subtypes, which differ across multiple defining variables; however, differences in cigarette smoking and cessation between these subtypes have not been previously investigated. Topiramate reduces heavy drinking and has separately been found to help non-alcoholdependent individuals quit smoking. This study tested the hypothesis that topiramate's effects on smoking would be moderated by alcohol-dependence subtype, and explored craving as a mediator of this response. Method: One hundred twenty-nine abstinent alcohol-dependent outpatient male smokers participated in this 12-week, randomized controlled trial comparing topiramate (maximum dosage 200 mg/day) with placebo, both with brief counseling, for smoking cessation. Participants were followed for 24 weeks following end of treatment. Results: Of the 125 participants with sufficient subtyping data, k-means cluster analysis categorized 52 (42%) as Type A alcoholics and 73 (58%) as Type B. Types A and B did not differ on baseline smoking characteristics, urges to smoke, or smoking consequence scores. Longitudinal mixed-effects regression indicated that the effect of treatment on smoking was moderated by the Type × Time interaction. Specifically, during the nontreatment follow-up phase, Type B's treated with topiramate had relative suppressed levels of smoking compared with placebo-treated Type B's. This moderating effect of the Type × Time interaction was mediated by intention to smoke and craving related to relief of negative affect. Conclusions: Type B alcoholics demonstrated suppressed levels of smoking in response to topiramate treatment as compared with placebo, but only during the nontreatment follow-up phase. This effect was mediated, in part, through intention to smoke and craving to smoke to relieve negative affect. Our findings extend other studies demonstrating a differential medication response by alcoholism subtype. (J. Stud. Alcohol Drugs, 78, 232-240, 2017)