The principal objective of this five-year longitudinal study was to determine the efficacy of employing the Omnibus Personality Inventory (OPI) and the American College Testing (ACT) program scores in designing a predictive model for identifying entering community college freshmen students' propensity to fall into one of three categories: successful completers, successful withdrawers (those who withdraw in good academic standing), and unsuccessful withdrawers. The study traced the academic performance patterns and achievement of 1,393 first-time day students in a California community college during their college stay, compared students' non-intellective characteristics to demonstrate the degree and dimension of differences between the three achievement groups, and analyzed characteristics that might be included in a predictive model. Comparisons between groups were made in terms of units attempted and completed, academic achievement and grade point average (GPA), academic probation and qualification data, number of terms attended, ACT and OPI response patterns, and normative national samplings. Findings indicated that although there were clear differences in GPA and ACT scores between groups, academic performance and persistence of the three groups was not accounted for by the test scores used in the study, and no one variable or combination of variables employed accounted for academic persistence. (TR)