All but seven U.S. states have laws that govern some aspects of dyslexia screening, intervention, or teacher training in public schools. However, in the three states that mandate child-level reporting, data indicate lower than expected rates of dyslexia identification when compared with commonly accepted dyslexia prevalence rates. To better understand this apparent mismatch, this study explores factors that might predict the school-assigned identification of individuals with dyslexia. Deidentified data on 7,947 second-grade students in 126 schools from one U.S. state included a universal screening measure of literacy skills indicative of dyslexia (i.e., reading and spelling), school-assigned dyslexia classification, and demographic characteristics. As expected, behavioral characteristics of dyslexia from universal screening were associated with school-assigned dyslexia classification. However, dyslexia classification was less likely for minority students and individuals attending schools with a higher percentage of minority students. Students who showed behavioral characteristics of dyslexia and attended schools with a higher proportion of other students with similar poor literacy skills were more likely not to receive a school-assigned dyslexia classification. The findings suggest systematic demographic differences in whether a student is identified with dyslexia by schools even when using universal screening.