Over one-third of college students in the United States transfer between institutions, yet little is known about how transferring affects students' educational and labor market outcomes. Using administrative data from Texas and a regression discontinuity design, I study the effects of a student's transferring to a four-year college from either a two-year or four-year college. To do so, I leverage applications and admissions data to uncover unpublished GPA cutoffs used for transfer student admissions at each institution and then use these cutoffs as an instrument for transfer. In contrast to past work focused on first-time-in-college students, I do not find positive earnings returns for academically marginal students who transfer from two-year colleges to four-year colleges or from less-resourced four-year colleges to flagship colleges, and show suggestive evidence of negative returns. The mechanisms include transfer students' substitution out of high-paying majors into lower-paying majors, reduced employment and labor market experience, decreases in academic performance relative to college peers, and potential loss of support networks. [Additional funding provided by the Dorothy Rice Dissertation Fellowship and the Mary Claire Ashenbrenner Phipps Dissertator Fellowship.]