We describe results of a long-term design experiment focused on promoting mathematics learning among multilingual ninth graders classified as English Learners. The intervention at a linguistically diverse public high school in the US focused on a unit introducing concepts related to linear rates of change. We analyze results from a curriculum-aligned pre- and post-unit assessment used to document student learning across each design cycle. The main result is that students in both the Pre-Intervention and Redesigned classrooms made gains on the pre- and post-unit assessments. However, student gains on the assessments were higher in the Redesigned classrooms than in the Pre-Intervention classrooms. Additionally, on the assessment in the Redesigned classes, students classified as English Learners made larger gains than their non-EL peers, and the majority of the gains occurred on conceptually-focused items. [For the complete proceedings, see ED629884.]