Confronted with many challenges to improving the quality of U.S. public K-12 education, many policymakers have viewed the study of a second language as a useful but nonessential ingredient of a world-class education system. However, others point out that dual-language education can be a powerful intervention for closing the achievement gap for English language learners (ELLs), and that it enhances outcomes for both ELLs and English native speakers. Though numerous studies have established a positive relationship between dual-language immersion (DLI) and student achievement in core subjects like mathematics and science, questions about these direct relationships remain. The proposed paper addresses three research questions about DLI: (1) What is the effect of dual-language immersion on student achievement in math, science; English language arts, and the partner language, and on student attendance and behavior? (2) How do the achievement, attendance, and behavioral effects of dual-language immersion differ for native English speakers versus native speakers of the partner language?; and (3) How does DLI program implementation vary across programs, and does that variation illuminate differences in program effectiveness? The study takes place in the Portland Public School district (PPS) in Portland, Oregon, the largest school district in the Pacific Northwest. PPS serves approximately 47,000 students in grades pre-kindergarten through 12. The study focuses in particular on the district's 19 dual-language immersion schools, which include 14 schools with Spanish programs, 3 with Japanese programs, 3 with Russian programs, and 2 with Mandarin programs. Analysis of the first years of data from a 3-year study are still under way. Preliminary descriptive analyses point to the importance of the random assignment strategy in removing selection bias. One figure is appended.