College students frequently engage in drinking games (DGs) and experience a variety of consequences as a result. It is currently unknown whether brief motivational interventions (BMIs) that provide feedback on DG participation can reduce this high risk behavior. This study examined outcome data from 2 randomized clinical trials to examine whether BMIs facilitate change in DG frequency and how these changes may occur. Mandated college students (Trial 1, N = 198, 46% female; Trial 2, N = 412; 32% female) were randomized to BMI or comparison control conditions. Hierarchical linear modeling (HLM) was used to compare the BMI and comparison groups to determine whether the BMI reduced DG participation over time. Percent change talk (PCT) during the discussion of DG during the session was examined as a predictor of change in DG frequency, and gender was examined as a moderator of treatment effects. Controlling for regular drinking frequency, participants who received a BMI did not significantly reduce their DG frequency relative to the comparison group in either sample, and the BMI was equally ineffective at reducing DG behavior for men and women. DG-related PCT during the BMI was associated with lower DG frequency at the second follow-up in both trials. In Trial 1, PCT during the BMI was associated with less steep increases in DG frequency across the course of all follow-ups. Effects of PCT on DG behavior were not moderated by gender. Findings did not support hypothesized reductions in DG participation following a BMI. Future research should explore whether targeted DG-specific interventions could reduce DG participation and the role of in-session client language in facilitating such change.