Glossary and reading aloud test items are commonly allowed in many states' accommodation policies for English language learner (ELL) students for large-scale mathematics assessments. However, little research is available regarding the effects of these accommodations on ELL students' performance. Further, no research exists that examines how students use the accommodations. Using a randomized experimental design and think-aloud procedure, the present study explored the effects of the two accommodations. Results showed no significant effect of glossary, mixed effects of read aloud, and significant interaction effects between students' prior content knowledge and accommodations, which suggests the accommodation was effective for students with content knowledge. Think-aloud data indicated that students were not familiar with, and did not actively utilize the glossary. Implications for the effective use of accommodations and future research are discussed. (Contains 1 footnote, 11 tables, and 2 figures.)