The authors conducted two randomized clinical trials with ethnically diverse samples of college student drinkers in order to determine (a) the relative efficacy of two popular computerized interventions versus a more comprehensive motivational interview approach (BASICS) and (b) the mechanisms of change associated with these interventions. In Study 1, heavy drinking participants recruited from a student health center (N = 74, 59% women, 23% African American) were randomly assigned to receive BASICS or the Alcohol 101 CD-ROM program. BASICS was associated with greater post-session motivation to change and self-ideal and normative discrepancy relative to Alcohol 101, but there were no group differences in the primary drinking outcomes at 1-month follow-up. Pre to post session increases in motivation predicted lower follow-up drinking across both conditions. In Study 2, heavy drinking freshman recruited from a core university course (N = 133, 50% women, 30% African American) were randomly assigned to BASICS, a web-based feedback program (e-CHUG), or assessment-only. BASICS was associated with greater post-session self-ideal discrepancy than e-CHUG, but there were no differences in motivation or normative discrepancy. There was a significant treatment effect on typical weekly and heavy drinking, with participants in BASICS reporting significantly lower follow-up drinking relative to assessment only participants. In Study 2, change in the motivation or discrepancy did not predict drinking outcomes. Across both studies, African American students assigned to BASICS reported medium effect size reductions in drinking whereas African American students assigned to Alcohol 101, e-CHUG, or assessment did not reduce their drinking.