ABSTRACT The present study performed a systematic comparison of DSM-5 and ICD-11 diagnostic criteria for reading disability. We quantitatively investigated the consequences of using DSM-5 or ICD-11, and of the different ways of implementing each diagnostic criterion on the prevalence of reading disability. We did so in a representative sample of the population of French sixth-graders (N = 25,000), using a reading comprehension test to assess reading ability. A compromise set of criteria and thresholds yielded a prevalence of 6.6% according to DSM-5 and 3.5% according to ICD-11. Factors that had the greatest influence on prevalence estimates were the criteria relative to IQ and to interference with academic performance. Compared with the reference population, children with reading disability were more likely to be boys (sex ratio≈1.6), to be schooled in a disadvantaged area (OR≈2.1), and to have lower SES (d≈-0.7), non-verbal IQ (d≈-0.4 – -0.9), and math scores (d≈-1.4). Our results emphasize that the choice of classification and the operationalization of diagnostic criteria have a large impact on who is diagnosed with reading disability.