This study of prediction of college success was designed to use the high school GPA and other variables, both cognitive and non-cognitive. It attempted to locate factors that would predict grades in English, mathematics, and biology, as well as predict the over-all GPA. Seven hundred students were examined. In college English, 49% of the variance could be determined from the student's high school English average, his high school GPA, and his score on the Brown-Holtzman Survey of Study Habits and Attitudes. Grades in mathematics were predicted from high school mathematics average, high school GPA, and the Florida 12th-Grade Mathematics Aptitude score. The best single predictor of success in college English and mathematics was high school success in them. Predictions for success in biology were based on the high school biology grade, the high school GPA, the Florida 12th-Grade Natural Science Score, and the Edwards Personal Preference Abasement Score. The best predictor in this course, however, was the over-all high school GPA. In summary, the multiple regression study of 31 variables generally supports the efficacy of considering past grades to predict grade success at the next level. It has also resulted in substantial multiple R's and has introduced non-cognitive variables not previously studied with any frequency. (HH)