We study the impact of a personalized mentoring program on university enrollment choices and academic outcomes. Conducting a randomized controlled trial among 337 high school students, we find that the program significantly influences students' decisions. The likelihood of choosing a field aligned with their mentor increases by 14 to 22 percentage points, depending on the sample and specification, representing a 25% to 45% increase from the baseline. Notably, the program also shifts preferences towards STEM/Economics fields, potentially enhancing prospective wages by 3.1- 3.7%. Using administrative data, we confirm the validity of survey-based evidence and show that the intervention does not negatively impact university performance, even though treated students enroll in more competitive fields.