College and major choices are life-altering decisions with significant long-term labor market consequences. However, searching and learning about all available programs are extremely costly, and the application decisions are further complicated when the centralized matching mechanism is not strategy-proof. In this project, we conduct information interventions among high school students in China. We first survey the students about their preferences towards different types of colleges and major categories and measure their risk preferences, which enables us to provide personalized information intervention. In the randomized trial, we recommend college-major programs based on the treated students’ revealed preferences, so that they become better informed about the availability of suitable programs without the costly search. In addition, we tailor our recommendations by incorporating the predicted probability of admission to each of the programs, improving the sophistication of the students’ rank-order lists. The effectiveness of the personalized recommendation is evaluated based on students’ submitted rank-order lists, actual admission results, and their levels of satisfaction. We also explore the general equilibrium effects by conducting a simulation analysis assuming the personalized recommendation is available to all students.