In general, need-based financial aid positively affects students’ academic outcomes and can reduce the amount students borrow. Evaluations of the federal Pell Grant program, however, have found mixed results. In this paper, we investigate possible explanations for often small or null effects of the Pell Grant documented in other studies using administrative data from Kentucky. We focus on whether and how (1) year-to-year changes in aid eligibility and (2) interactions with other sources of grant and loan aid attenuate the Pell Grant’s effect on students’ educational attainment. Our work complements other evaluations of the Pell Grant program that use cross-sectional variation in aid eligibility by modeling the evolution of Pell Grant eligibility over the course of a students' college enrollment. We rule out year-to-year changes in Pell eligibility as an explanation for Pell's mixed track record in past evaluations. We also find some evidence of negative impacts on borrowing, in line with other recent studies investigating student borrowing behavior.