Aim: To estimate the prevalence of major birth defects among children with autism, the prevalence of autism in children with birth defects, and the risk for autism associated with having birth defects. Method: Retrospective cohort including all children born in Atlanta, GA, USA, 1986 to 1993, who survived to age 3 years and were identified through Georgia vital records. Children with autism and other developmental disabilities residing in Atlanta at ages 3 to 10 years in 1996 were identified through the Metropolitan Atlanta Developmental Disabilities Surveillance Program. Children with major birth defects through age 6 years were identified by the Metropolitan Atlanta Congenital Defects Program. Results: Birth defects were found among 6% of children with autism (total n=617; 488 males, 129 females) and was associated with a near twofold increased risk for autism overall. However, the risk magnitude and statistical significance varied by type of birth defect. With any type of birth defect, the risk for autism accompanied by intellectual disability or other developmental disabilities was typically higher than the risk for autism alone. A 6:1 to 8:1 male bias was observed among children with autism and a birth defect. Interpretation: Investigation of the association between autism and birth defects is warranted, especially for the role of birth defects in autism among sex-specific or autism subgroups.