English Learners, students who are not proficient in English and speak a non-English language at home, make up more than 10 percent of the nation's K-12 student body. Achieving proficiency in English for these students is a major goal of both state and federal education policy, motivating the provision of bilingual education policies. Using data for nearly 500,000 English Learners from California, I show that students in bilingual education have substantially lower English proficiency than other English Learners in first and second grades. In contrast, there is little difference between bilingual education and other programs for students in grades three through five. These results hold across fixed effects, propensity score, and instrumental variables models. (Four appendixes are included: (1) OLS (Ordinary Least Squares) Results; (2) Instrumental Variables First Stage Results--Instruments are Measures of Proposition-227 Induced Changes in School-Level Bilingual Education Availability; (3) Instrumental Variables First Stage Results--Instrument is the Percentage of "Yes" Votes on Proposition 227; and (4) Bilingual Education Interacted with Previous Listening/Speaking Quartile Dependent Variable is Reading or Writing Proficiency for Each Grade and Year. Contains 15 tables, 1 figure, and 20 footnotes.)