Research support for providing a read-aloud accommodation (i.e., having an individual read test items and directions aloud) to students with disabilities has been somewhat limited, particularly when merely examining effects of the accommodation on overall test scores for general groups of students with disabilities. We examined data on accommodated and nonaccommodated performances of students with specific reading disabilities on math test items anticipated to be highly sensitive to accommodation effects due to their level of reading and math difficulty. We conducted analyses across 3 consecutive years of data from an elementary and middle school statewide assessment program. Within the elementary data set we identified a significant Accommodation Group x Reading Difficulty interaction effect, such that the accommodation appeared to more positively impact student performance on items that were classified as difficult to read. We describe other identified effects and discuss limitations of the study, as well as implications. (Contains 5 tables, 2 figures, and 1 note.)