Low-income students are underrepresented at selective colleges. Most evaluative criteria used by admissions officers are correlated with income, including test scores and performance in high school. Inspired in part by the current state of the use of testing in college admissions, this study examines how the quality of colleges attended by students with varying levels of academic achievement and income changes when colleges’ testing policies change. I employ agent-based simulation – a computational technique used to study systems of emergent behavior – to build a model of the college choice process, incorporating both student and college decisions. I examine how students sort to colleges under a variety of different model conditions and testing regimes. This study finds that requiring tests and using test-optional admissions review both result in income-influenced mechanisms by which students sort to colleges. Test-required admissions enables low-income students with high test scores to sort to selective colleges at higher rates than test-optional admissions; however, low-income students with lower test scores are disadvantaged. While low-income students with lower test scores and higher high school performance enroll at higher quality institutions under test-optional admissions review, the practice benefits high-income students with lower test scores to a greater extent. As such, test-optional admissions review is an imprecise and only mildly effective mechanism for increasing access to selective colleges for low-income students. Because test-optional admissions review is not a panacea for recruiting low-income populations, enrollment management professionals must be deliberate in designing and emphasizing income-neutral criteria in the admissions process. Further, this study elucidates how a change in testing policy at one college has ripple effects on other colleges’ policy decisions. This study opens avenues for future research using agent-based models of college choice to explore other phenomena, including pricing behavior on the part of colleges, the impact of demographic change and demand shifts on the marketplace, and the impact of policy change on the market for higher education.