This article reports on a study that investigated achievement in math of third and fourth grade dual language immersion (DLI) students, building on research that has demonstrated the academic achievement of students who receive content instruction predominantly in the target language. Our study expands the scope and methodology of prior research by including one-way programs in three languages (Chinese, French and Spanish) and two-way Spanish-English programs; and by relying on propensity matching to mitigate possible effects of school and student differences. In our third grade study, we compared students' math scores in relation to their English Language Arts (ELA) achievement to control for pre-existing differences between DLI and non-DLI students. DLI students who attained the same levels in ELA, and who received math instruction in a target language, performed at the same level as their non-DLI peers in third grade math tests given in English. For the fourth grade study, we compared DLI students to a propensity-matched non-DLI group. DLI students grew more in math than their counterparts not in DLI. The results from this natural experiment indicate that students in a DLI program that has been implemented state-wide were able to succeed academically in math.