In this study, we use microdata from 12 Florida county-level school districts and a regression discontinuity design to examine the effects of early grade retention on the short-, medium-, and long-term outcomes of English learners. We find that retention in the third-grade substantially improves the English skills of these students, reducing the time to proficiency by half and decreasing the likelihood of taking a remedial English course in middle school by one-third. Grade retention also roughly doubles the likelihood of taking an advanced course in math and science in middle school, and more than triples the likelihood of taking college credit-bearing courses in high school for English learners. We also find that these benefits are larger for foreign born students, students with higher latent human capital in third grade as proxied by their math scores, students whose first language is Spanish, and students in lower-poverty elementary schools.