Data mining is increasingly becoming an essential tool in organizations today. Particularly, academic organizations are requesting more sophisticated tools to improve their decision making process. A large quantity of data and information is produced during the student’s life, but it is still necessary to turn them into insight. This paper describes a project that use data mining to support the decision making process in higher education in Chile. The aim of this project is to find patterns that allow the identification and determination of relationships among the initial conditions of students and with their final status as a student (drop-out or graduated). The study is conducted in a university in the north of Chile and it considers five undergraduate majors. The final results of this project are expected to support the decisions making process related with university admission policies, causes of student failure or success, and university marketing policies.