Using institutional data, this paper presents a model to enable researchers and enrollment managers to assess the effectiveness of financial aid packaging policies in light of student characteristics and institutional market position. The model uses discriminant analysis and a series of hypothetical financial aid award scenarios to predict the enrollment decisions of students who applied to an urban, comprehensive research institution with selective admissions criteria in autumn 2001. The findings suggest that changes in the need-based award policy would result in relatively small changes in enrollment yield. (Contains 5 figures.)