This paper recommends new strategies in the use of data in monitoring enrollment distribution, retention, and graduation rates of minority students in predominantly white colleges. Surveys of current literature show there are three significant questions: (1) Who drops out? (2) Why do they drop out? and (3) How can the overall dropout rate be reduced? The Data-Driven Minority Student Retention Model addresses these questions by utilizing a comprehensive approach that provides for use of data in improving retention and graduation rates of minority populations. The data collection and formating component of the model establishes a comprehensive and accurate data retrieval system for determining application, acceptance, and admission rates; academic strategy profiles; classification profiles; longitudinal program profiles; and graduation rates. The Computerized Academic Monitoring System component is the major force behind the early identification of students with academic difficulties during the first third of the term in selected courses. This system provides for the instructors' identification of students with poor marks and a follow-up communication system for contacting and assisting students. The adoption and effective implementation of the Data-Driven Retention Model should significantly improve the retention and graduation rates of minority students in higher education institutions. (Author/MK)