Early onset in adolescent gambling involvement can be a precipitator of later gambling problems. The aim of the present study was to test the preliminary efficacy of a web-based gambling intervention program for students within a high school-based setting. Students attending a high school in Italy (N = 168) participated in the present study (58% male - age, M = 15.01; SD = 0.60). Twelve classes were randomly assigned to one of two conditions: intervention (N = 6; 95 students) and control group (N = 6; 73 students). Both groups received personalized feedback and then the intervention group received online training (interactive activities) for three weeks. At a two-month follow-up, students in the intervention group reported a reduction in gambling problems relative to those in the control group. However, there were no differences in gambling frequency, gambling expenditure, and attitudes toward the profitability of gambling between the two groups. In addition, frequent gamblers (i.e., those that gambled at least once a week at baseline) showed reductions in gambling problems and gambling frequency post-intervention. Frequent gamblers that only received personalized feedback showed significantly less realistic attitudes toward the profitability of gambling post-intervention. The present study is the first controlled study to test the preliminary efficacy of a web-based gambling intervention program for students within a high school-based setting. The results indicate that a brief web-based intervention delivered in the school setting may be a potentially promising strategy for a low-threshold, low-cost, preventive tool for at-risk gambling high school students.