Population-based disease registry systems are invaluable research resources due to their non-biased coverage, prospective data collection, large size, and length of follow up. The relative rarity of autism, as well as many perinatal conditions, makes use of large multi-national registry-based samples an ideal approach for filling important gaps in our understanding of the relation between perinatal factors and autism. This proposal builds on the Autism Speaks-CDC International Autism Epidemiology Network Registry Workgroup. The Workgroup established the International Collaboration for Autism Registry Epidemiology (iCARE) and identified relevant registry systems for the proposed investigation in Scandinavia (Denmark, Sweden, Finland and Norway), Australia and Israel. The goals of the investigation are 1) to establish the necessary infrastructure for multi-registry autism research, including establishing a multi-registry virtual data set approach to analysis and establishing written guidelines for multi-national registry-based research collaboration, and 2) demonstrate the capabilities of the multi-registry approach to elucidate etiological pathways to autism by thoroughly investigating candidate factors for perinatal adversity in relation to autism, beginning with gestational length and fetal growth. The study will consider variation in diagnostic criteria and trends over time, and based on heretofore unrivaled sample sizes that will permit examination of risk across more finely defined birth weight and gestational age categories; consider effects on risk of fetal size at a given gestational age; control for confounding; and conduct robust stratified analyses considering other important risk factors, such as gender, parental age, and birth multiplicity, and features of autism, such as diagnostic subtype. The proposed project presets a new scope in epidemiological investigation into risk factors for autism, and will provide the foundation and guidelines for future investigations.