This paper examines the effect of a short-lived increase in tuition rates on the persistence and grades of undocumented college students.  In the spring of 2002, the City University of New York (CUNY) reversed its policy of charging in-state tuition rates to undocumented college students who could demonstrate that they finished their secondary schooling in New York.  With a full course load of 12 credits, this policy change represented a tuition increase of $1800 for bachelor’s degree seeking undocumented students.  The new policy was in place for exactly one semester (Spring 2002) and was subsequently overturned by the New York State legislature, which passed a law in the summer of 2002 that reinstated in-state tuition benefits for eligible undocumented students.  Using this unusual policy context and a unique dataset that records students’ immigration status, this paper estimates the impact of tuition increases on undocumented students’ persistence, full-time (versus part-time) enrollment, and grades.  The estimation strategy compares the difference between undocumented and documented students college outcomes in the tuition hike semester to the differences between these two groups in outcomes in nearby semesters. The results suggest that the tuition increase caused an 8% decrease in persistence and a 6% decrease in full-time enrollment for undocumented BA degree seeking students in the Spring of 2002.  No impact is observed on grades.  Further, though the tuition increase caused students to disenroll, the return of the subsidy in the next semester induced them back into school, indicating that the short-term tuition shock did not cause students to permanently drop out of college.  The findings provide strong evidence that college costs can have a large impact on the collegiate outcomes of undocumented students who have already chosen to attend college and, perhaps more generally, the outcomes of price sensitive college students, including those from low-income backgrounds.