Disproportionate representation of racially diverse students in special education is a well-documented problem, yet few studies have systematically evaluated disproportionate representation of diverse students with autism. This study examined disproportionate representation of racially diverse students with autism by determining risk and logistical odds ratios among racially diverse and White students from the national population between 1998 and 2006. Although overall risk of autism increased for all racial groups every year, White students were twice as likely to be identified with autism as Hispanic and American Indian/Alaskan Native students during most years of the sample. Although initially overrepresented, the odds ratios for Asian/Pacific Islander and Black students with autism continuously declined in recent years. Hispanic and American Indian/Alaskan Native students were significantly underrepresented every year in the analysis. Potential cause and implications of underrepresentation are described, along with directions for research. (Contains 1 table and 1 figure.)