Alcohol consumption has been a growing concern at U.S. colleges, particularly among first-year students, who are at increased risk for problems. This study tested the efficacy of the "electronic Check-Up to Go" (e-CHUG), a commercially-available internet program, at reducing drinking among a group of at-risk college freshman. The design was a randomized controlled trial: 106 freshmen students who reported heavy episodic drinking were randomly assigned to receive feedback or to assessment only. Assessment measures were completed at baseline, 8 weeks, and 16 weeks. At 8 weeks, the feedback group showed a significant decrease in drinks per week and peak BAC over control. By 16 weeks, the control group also declined to a point where there were no differences between groups. Changes in normative drinking estimates mediated the effect of the intervention. An additional 245 abstainers and light drinkers who were also randomized to condition did not show any intervention effect. This study provides preliminary support for the efficacy of this intervention at reducing short-term drinking among at-risk students.