OBJECTIVES: We sought to examine racial and ethnic disparities in the recognition of autism spectrum disorders (ASDs). METHODS: Within a multisite network, 2568 children aged 8 years were identified as meeting surveillance criteria for ASD through abstraction of evaluation records from multiple sources. Through logistic regression with random effects for site, we estimated the association between race/ethnicity and documented ASD, adjusting for gender, IQ, birthweight, and maternal education. RESULTS: Fifty-eight percent of children had a documented autism spectrum disorder. In adjusted analyses, children who were Black (odds ratio [OR] = 0.79; 95% confidence interval [CI] = 0.64, 0.96), Hispanic (OR = 0.76; CI = 0.56, 0.99), or of other race/ethnicity (OR = 0.65; CI = 0.43, 0.97) were less likely than were White children to have a documented ASD. This disparity persisted for Black children, regardless of IQ, and was concentrated for children of other ethnicities when IQ was lower than 70. CONCLUSIONS: Significant racial/ethnic disparities exist in the recognition of ASD. For some children in some racial/ethnic groups, the presence of intellectual disability may affect professionals' further assessment of developmental delay. Our findings suggest the need for continued professional education related to the heterogeneity of the presentation of ASD.