We conducted a best-evidence synthesis of 22 studies to examine whether systemic bias explained minority disproportionate overrepresentation in special education. Of the total regression model estimates, only 7/168 (4.2%), 14/208 (6.7%), 2/37 (5.4%), and 6/91 (6.6%) indicated statistically significant overrepresentation for Hispanic, Asian, Native American, and English language learner (ELL) or language-minority children, respectively. Among studies with the strongest internal and external validity, none of the 90 estimates (i.e., 0%) indicated overrepresentation attributable to racial or ethnic bias. Of the 18 estimates for language-minority and ELL children combined, only 3 (16.7%) indicated overrepresentation attributable to language use. Two of the 4 ELL-specific estimates (50%) indicated that children receiving English-as-a-second-language services may be overrepresented in special education. Overall, and replicating findings from a prior best-evidence synthesis, this synthesis indicated that children are underidentified as having disabilities based on their race or ethnicity and language use.