Objective: This study tested mediating processes hypothesized to explain the therapeutic benefit of an efficacious motivational interview (MI). The constructs of interest were motivation to change, cognitive dissonance about current drinking, self-efficacy for change, perceived young adult drinking norms, future drinking intentions, and the use of protective behavioral strategies. Method: A randomized controlled trial compared the efficacy of a brief MI to a time- and attention-matched control of meditation and relaxation training for alcohol use. Participants were underage, past-month heavy drinkers recruited from community (i.e., non 4-year college or university) settings (N = 167; ages 17–20; 58% female; 61% White). Statistical analyses assessed mechanisms of MI effects on follow up (6-week, 3-month) percent heavy drinking days (HDD) and alcohol consequences (AC) with a series of temporally lagged mediation models. Results: MI efficacy for reducing 6-week HDD was mediated by baseline to postsession changes in the following 3 processes: increasing motivation and self-efficacy, and decreasing the amount these young adults intended to drink in the future. For 6-week AC, MI efficacy was mediated through 1 process: decreased perceived drinking norms. At 3-month follow up, increased cognitive dissonance mediated HDD, but not AC. Further, increased use of certain protective behavioral strategies (i.e., avoidance of and seeking alternatives to drinking contexts) from baseline to 6-weeks mediated both 3-month HDD and AC. Conclusions: Findings suggest that within-session cognitive changes are key mechanisms of MIs effect on short-term alcohol outcomes among community young adults while protective behaviors may be more operative at subsequent follow up.