The link between reading and mathematics achievement is well known, and an important question is whether readability factors in mathematics problems are differentially impacting student groups. Using 20 years of data from the National Assessment of Educational Progress and the Trends in International Mathematics and Science Study, we examine how readability factors--such as length, word difficulty, and pronouns--interact with student background characteristics--such as race/ethnicity, mathematics achievement, and socioeconomic status. Textual features that make problems more difficult to process appear to differentially negatively impact struggling students, while features that make language easier to process appear to differentially positively impact struggling students. It is critical that readability along various dimensions be considered when designing instruction and assessment.