The underidentification of students with autism is a well-documented phenomenon that implies minority children are less likely to receive autism-specific interventions and services vital for better outcomes. Previous research has found that minority children are less likely to be identified as having autism, but recent changes in federal reporting from five to seven racial categories warrant further investigation of minority underidentification in the autism category. This logistic regression analysis of autism count data was conducted to examine differences in identification according to seven racial reporting categories. We calculated the probability of autism identification for each racial group and tested odds ratios for significant differences within each state and Washington, DC. We also compared odds of autism identification for each racial group in each state to the odds of identification for White students in California. Within-state analyses indicated minority students in most states continue to be identified at rates significantly lower than their White counterparts. Between-state analyses revealed White students and minority students in most states were underidentified compared to White students with autism in California. The change to seven racial reporting categories appears to have clarified the extent to which minority students are underidentified in the autism category, which may mean minority students experience restricted access to evidence-based interventions and services for students with autism.