Researchers across the world are trying to understand precursors of autism spectrum disorders (ASD) and tools for the early identification of this disorder. Early identification is essential for enabling early intervention. Studies have shown that there are markers of ASD that can be detected in infancy among them deficits in joint attention and social interest. Available ASD screening tools have so far met with partial success in population screening thus cannot be employed in practice. Our goal is to study the predictive validity of ASD screening in infancy to later diagnosis, and the feasibility of ASD screening within the health care system. We will pursue our goal by conducting a two-stage screening study of a birth-cohort of 18-month old infants. At first infants will be screened with a parent questionnaire during their 18-month visit to baby wellness clinics. Screened negative infants will be further screened using a phone interview. Infants who continue to show risk for ASD will be administered a standardized diagnostic evaluation. A subset of children who screen negative will also be administered a diagnostic evaluation. This is crucial for identifying infants who were missed in the initial screening and for reducing bias of predetermined screening criteria. The efficacy of initial screening will be determined based on its prediction to diagnosis. Previous research by the PI, conducted mainly at Boston University and University of Massachusetts Boston, has shown that young children with ASD show sensory processing abnormalities early on. This proposal aims to expand upon previous research knowledge and experience with young children with ASD at the host institution. The host institution has world-class researchers in many of the disciplines to which early identification of ASD can be applied. We expect to interact and collaborate with researchers across the diverse fields at the host institution thus ensuring lasting reintegration.