This project is focused on the development, validation, and standardization of quantitative methods to measure core components of variation in the social development of young children. This is critical for understanding autism and will inform the development and evaluation of early interventions for children with autism and related conditions. The proposed application covers the 18 to 48 month age range; i.e. the period when autism is actually first diagnosed. Currently, there exist few options for the tracking of autisticseverity over time in this age range, few methods that allow severity trajectories to be reliably associated with candidate genetic and neurobiological correlates of autistic impairment, and few established methods for making meaningful baseline measurements against which to quantify the impact of early intervention for young children with autism spectrum disorders (ASD). In this application, as we attempt to extend to early childhood a measurement system rooted in the ascertainment of quantitative autistic traits, we couple such measurements with laboratory assessment of visual social engagement (VSE), a fundamental parameter of human social development, whose origins are traceable to earliest infancy and whose implications are wide-ranging with respect to all aspects of early human social development. The specific aims are as follows: 1) To examine the general population distribution and longitudinal course of quantitative variation in reciprocal social behavior (and related indices of social behavioral development) in an ethnically diverse sample of preschool age twins (n=420 pairs), assessed prospectively at key ages within the interval from 18-48 months; 2) To examine stability and change in the same set of quantitative measures when implemented among children with ASD (n=90) first diagnosed at 18-24 months, and to clarify the relationship between these measures and traditional diagnostic assessments for ASD; 3) To examine variation in visual social engagement (VSE)-a neurodevelopment capacity compromised in children with ASD-among preschool children (125 general population twin pairs and 75 clinical ASD subjects) representing the entire range of social variation that occurs in nature.