This study will address the function of young adults with autism spectrum disorders (ASDs) and the child and adolescent factors that predict good versus poor outcome. A well-characterized, longitudinal and population-representative cohort of young adults with ASD will be studied to answer the following questions: 1) What is the range of outcomes for young adults?; 2) What are the factors, especially from childhood and adolescence, that predict good vs. poor outcome?; 3) What are the drivers of the high economic costs of ASD in adulthood, and to what extent can they be identified in childhood?; and 4) Can we reliably identify cognitive markers for associated psychiatric problems that may help in earlier diagnosis and treatment of these additional problems? The ultimate aim is to identify features that point to novel ways of improving outcome by preventing and/or treating additional impairments. This unique cohort, the Special Needs and Autism Project (SNAP), is derived from the general population rather than through clinics. Therefore, it represents the entire spectrum with respect to autism severity, IQ and family characteristics such as education and ethnicity. Findings from this study will represent all young adults with ASD, not just those accessing services. The cohort has been previously assessed at ages 11/12 years and 15/16 years and the data include measures of autism, global and specific cognitive function, co-occurring psychopathology and family and contextual characteristics. The investigators will examine a range of adult outcomes, including medical and mental health problems, adaptive functioning including employment, independent living and use of social care services, quality of life, and parental burden and well-being. They will identify the personal, family and broader social factors, both in childhood and adolescence, which predict adult outcome. However, the study focuses particularly on the role of co-occurring psychopathology and hypothesizes that this will be a strong, independent predictor of poor adult outcome, alongside established predictors such as language and IQ. Furthermore, it is predicted that many of the psychiatric problems seen in adults with ASD have been present since childhood and stable over time. If correct, the study will have identified tractable targets for future studies on early diagnosis and treatment of co-occurring psychopathology to improve adult outcome.