The rates of Cesarean delivery (C-section) have risen to >30 % in numerous countries. Increased risk of autism has been shown in neonates delivered by C-section. This study examined the incidence of autism in neonates delivered vaginally, by C-section with regional anesthesia (RA), and by C-section with general anesthesia (GA) to evaluate the risk of autism associated with C-section and obstetric anesthesia. During a mean follow-up of 4.3 years, the incidence of autism was higher in neonates delivered by C-section with GA than in neonates delivered vaginally, with an adjusted risk of 1.52 (95 % confidence interval 1.18–1.94). However, the adjusted risk of autism in neonates delivered by C-section with RA and in neonates delivered vaginally was nonsignificantly different.