A recurring issue among community colleges is the perceived high rate of attrition of their students. Summers (2000) conducted a study at a Midwest comprehensive community college to examine the relationship between student enrollment behaviors and various outcomes. He found that students who persisted enrolled earlier than those who did not. He also found that for each additional course dropped the likelihood of enrolling the following spring decreased by 50%. Based in part on this study some community colleges have begun to limit full matriculation after a given date at the beginning of the fall semester allowing late-start students to enroll part-time in "late-start" classes. The purpose of this study is the replicate Summer's findings using a similar cohort but at a different community college and then to apply the same techniques, if possible, to "late start" classes. Phase One of this study had similar findings to that reported by Summer. However, significant problems were encountered when similar techniques were applied to "late-start" classes. A brief review of relevant literature is included along with recommendations for further research. (Contains 14 tables.)