The study we report here examined parental, pre-course, and in-course predictors of students' probability of retaining (n = 136) or losing the HOPE scholarship (n = 41). The study was conducted in a multi-section, entry-level course (n = 203) for the Teacher-Education Program at a large state university in the southeastern U.S. Logistic regression that included a special computation of standardized beta weights was used in determining the predictive strength of selected variables and groups of variables. The variables with the strongest standardized beta weights included class attendance, critical thinking ability, and high school GPA. Of the various sub-groups of variables, pre-course and in-course variables generated the strongest prediction models. All models did much better in predicting retention than loss of HOPE scholarships.