Employing longitudinal, multi-site comparative mixed methods, we describe patterns of reclassifying and not reclassifying eligible English learner students to fluent English proficient, and we identify factors impeding and facilitating reclassification. Analyses of administrative data for 7 cohrts of students in 2 districts revealed a considerable number of students meeting all criteria, and a subset meeting at least standardized-test criteria, were not reclassified, implicating this practice in the production of long-term EL status. The rate of not reclassifying was more than 2 to 5 times higher in one district than the other. Analyses of policy documents, reclassification decision forms, staff interviews, and surveys, revealed that undergirding these differences were distinct reclassification policies and practices. We discuss policy, practice, and research implications.