Most studies have reported autistic children as coming from upper social economic status (SES) families. A few studies have not found any such social class bias. In order to resolve these contradictory findings, an empirical study was conducted on a statewide sample of families with autistic children. It was hypothesized that autistic children from high SES families would be associated with seven social class selection factors: (1) early age of onset, (2) early age of treatment admission, (3) normal cognitive potential, (4) complex rituals with maintenance of sameness, (5) long distance traveled for treatment, (6) limited availability of services, and (7) very detailed child history. Factors 1, 5, 6, and 7 distinguished high SES from low SES families in the predicted direction. Implications for research and treatment are discussed.