Using data from a nationwide longitudinal study of student attendance patterns in higher education, this study compared the responses of community college students who discontinued their enrollment during their initial semester with those of community college students who persisted for two years. Data for the analysis and comparison were opinion, attitude, belief, and socio-economic information. In addition, student scores on scales of the Omnibus Personality Inventory were compared. The two major hypotheses studied were: (1) that no pattern of scores on standard instruments and supplemental biographical questionnaires discriminates between students persisting and withdrawing in a national sample of community college students and (2) that discriminant scores contrasting a national sample of community college students who persisted for less than one semester with students who persisted for two years cannot be applied with acceptable empirical validity to an independent sample of students in two metropolitan community colleges. To test the hypotheses, discriminant scores were developed and an empirical validity was found for each hypothesis. Limitations of the model are discussed and recommendations for improving the prediction of individual attrition are made. In addition, strategies for approaching the potential dropout in the community college are suggested for further research. (Author/RC)