We use data from the 1990/94 Beginning Post-Secondary Survey to determine whether the factors associated with long-term attrition from higher education differ for students who initially enrolled part-time as compared to for students who initially enrolled full-time. Using a two-stage sequential decision model to analyze the initial enrollment intensity decision jointly with attrition, we find no evidence of correlation in the unobservables that necessitates joint estimation, but substantial evidence that the factors associated with attrition differ by initial enrollment status. The timing of initial enrollment, academic performance, parental education, household characteristics, and economic factors had a substantially greater impact on those initially enrolled full-time, while racial and ethnic characteristics had a greater impact on those initially enrolled part-time. The results of our study suggest that separate specifications are necessary to identify at-risk full-time as compared with at-risk part-time students.