In recent years, Strategic Enrollment Management (SEM) initiatives have become integral in the programs that colleges and universities develop to recruit and retain students. With 73.0% of full-time freshmen at Wichita State University in 2017 returning for their second year of study in fall of 2018, there is a need to implement interventions designed to identify and provide services to students at-risk of not being retained. Based on research of academic attrition and models of student retention, a conceptual model of student retention was developed. Wichita State University has developed collaborations across multiple SEM offices to increase their capacity to strategically use student record data to create data-driven programs and policies. This study utilized this capacity to develop a predictive model of student retention for three of the university's main student populations: first-time-in-college students, transfer students, and returning adult students. The availability of student data for each population is impacted by university admissions and data monitoring practices, requiring the conceptual model to be tailored to each student group. Bivariate comparisons between the students who were retained and who were not retained in each population revealed significant differences between the groups, so a logistic regression was used to predict retention risk. The logistic regression equations for each population were able to predict student retention with at least 70% accuracy. Implications, limitations, and suggestions for future research will be discussed.