We (Kranzler et al., 2014) reported that topiramate 200 mg/day reduced heavy drinking days and increased abstinent days in 138 heavy drinkers whose treatment goal was to reduce drinking to safe levels. In that 12-week, placebo-controlled study, we measured drinking using the Timeline Follow-back method at each treatment visit. In addition to the intent-to-treat effects of topiramate, we found that a single nucleotide polymorphism (rs2832407) in GRIK1, encoding the GluK1 subunit of the kainate receptor, moderated the treatment effect in European Americans (EAs; n = 122). Topiramate reduced heavy drinking only in rs2832407*C allele homozygotes. Here, we augment those analyses by using patients’ daily reports obtained using interactive voice response technology; (a) to validate the interactive effects of GRIK1 and topiramate as predictors of drinking level; and, (b) to examine changes in expected positive effects of drinking (i.e. positive outcome expectancies) and desire to drink. We found that rs2832407*C allele homozygotes treated with topiramate drank less overall during treatment than those receiving placebo, validating our earlier findings for heavy drinking days (Kranzler et al., 2014). There was also a study day × medication group × genotype group interaction that predicted both positive alcohol expectancies and desire to drink, with rs2832407*C-allele homozygotes treated with topiramate showing the largest decreases in these outcomes during the study period. Changes in positive alcohol expectancies or desire to drink did not mediate the effects on drinking. These findings validate and extend our previous pharmacogenetic findings with topiramate.