The incidence of autism in the Tuberous Sclerosis Complex (TSC) is markedly higher than that in the general population, with features of autism present in 25-60% of children with TSC. Clearly there exists a strong association between TSC and autism, and many research efforts have been directed towards elucidating the etiology of this association. There have been a number of retrospective studies evaluating markers of autism in TSC, but no studies have evaluated prospectively children with TSC for the development of autism. For the study of early signs and clinical predictors of autism, children with TSC serve as an ideal high-risk study population for two reasons. First, they are diagnosed early in life and are followed by neurologists throughout their childhood years. This facilitates longitudinal studies and minimizes attrition. Secondly, children with TSC are well characterized neurologically, with neuro-imaging, genetic testing, and electroencephalograms in early infancy as part of their baseline clinical evaluation. These data can promote recognition of phenotypes associated with autism. The study of autism in TSC may provide us with novel insights into the neuroanatomic and genetic correlates of autism. In this setting, we have created a prospective study using electrophysiological and behavioral assays to understand early markers and clinical correlates of infantile autism in children with TSC. The first aim of out study is to screen children with TSC for autism in infancy, using an event related potential (ERP) paradigm of face processing, and a behavioral test known as the Autism Observation Scale of Infancy (AOSI). We hypothesize that the behavioral abnormalities will be measurable in children with autism in the first year of life. The second aim is to prospectively analyze the genetic and neuroanatomic correlates of autism in TSC, with the goal of determining whether certain genetic mutations or specific patterns of brain injury predict an autistic phenotype. Specific variables to be evaluated include presence and location of intracerebral tubers and subependymal nodules, white matter tract abnormalities quantified with diffusion tensor imaging (DTI), EEG abnormalities, and TSC1 and TSC 2 mutations. Our hypothesis is that the most predictive variable for autism will be the presence of the TSC2 mutation. Also, DTI abnormalities may be more prominent in children with autism. The final aim of the study is to describe the specific behavioral phenotype of children with TSC, particularly in those children who meet criteria for autism. Is there a specific deficit in behavior that exists in all children with TSC, even those without a formal diagnosis of autism? Wehypothesize that the children with TSC will show more impairment in social interaction and play than in communication, and that there will be an inverse relationship between cognitive function and autism. What this means for people with autism:The study of autism in TSC may provide us with novel insights into the neuroanatomic and genetic correlates of autism. It also may help us to define markers of autism in early infancy in high-risk populations, which could then facilitate earlier interventions for these infants.