In recent years, epidemiologic studies have implicated a number of potential prenatal and perinatal risk factors in the etiology of autism. Unfortunately, challenges inherent in observational risk factor study designs, such as confounding, measurement error, and selection bias, can limit causal inference from epidemiologic studies of autism etiology. In this educational symposium, speakers will review the state of epidemiology research regarding four different prenatal risk factor domains noting the major methodological challenges specific to each domain and highlighting at least one specific technique that can be used to overcome these challenges.