Autism is a neurodevelopmental disorder that affects nearly 1 in 252 girls. Yet, little is currently known about the neurobiological basis of autism in girls, since extant studies have almost exclusively focused on boys or mixed gender samples involving a small number of girls. Autism is less prevalent in girls than boys, with a ratio of 1 to 4, raising the possibility that girls are protected from the disorder. In a series of investigations, I will characterize structural and functional brain organization underlying atypical cognition and behavior in girls with autism, using a novel 'big data' science approach that overcomes weaknesses associated with previously under-powered studies that fail to capture the underlying heterogeneity of the disorder. Specifically, I will collate, integrate, and analyze behavioral, cognitive, and brain imaging data collected at over 100 US and international sites and obtained from the National Institute of Mental Health - National Database of Autism Research and the Autism Brain Imaging Data Exchange. This will enable one of the most comprehensive brain-behavior investigations of autism heterogeneity, which I will undertake, using collated data from over 50,000 individuals with autism and typical controls, providing one of the largest cohorts of autism to date. The proposed use of a 'big data' science approach for translational research in psychiatry is first of its kind and poses unique set of challenges including, but not limited to, a) data complexity: brain data is inherently complex and provides information at multiple spatial and temporal resolutions, b) data heterogeneity: brain data is highly multimodal and usually acquired using different scanners and acquisition protocols at multiple sites, further complicating the development of field-wide consensus, c) large datasets: a typical dataset from one subject includes information from 106 brain voxels collected typically every 2s, and d) limited sample sizes: relative to the size of the data per subject the number of observations subjects) is very small, resulting in analytical challenges. With extensive research experience combined with my interdisciplinary training in quantitative methods and biomedicine, I am uniquely positioned to develop solutions that address these challenges. My research is critical not only for understanding the etiology of this heterogeneous disorder but also for understanding neuroprotective factors in girls. Importantly, given the need for a non-invasive gender-specific biomarker for autism, I will investigate whether quantitative measures of brain organization along with cognitive measures provide reliable biomarkers in girls with autism. My proposed research will provide novel gender-specific, systems-level insights into aberrant brain organization underlying cognitive and behavioral dysfunction in autism, and contribute significantly towards understanding the complex neurobiology of autism. Finally, the computational tools and 'big data' science approach developed here will be more generally applicable to other psychiatric disorders, including depression and schizophrenia.