ABSTRACT A major gap in our understanding of autism spectrum disorder (ASD) is in knowing how the brain functions during conditions that approximate the complex processing demands of the real world. Instead, almost everything we know about brain functioning in ASD comes from reductionist studies that often use highly simplified stimuli and isolated task demands, or even from studies that lack a stimulus or task altogether, as in the case of resting-state functional connectivity. Yet, successful processing “in the wild” (i.e., in the real world) relies on the simultaneous engagement and seamless integration of multiple brain regions, brain networks, and cognitive processes. Understanding how these neural systems behave and interact during real-time processing of complex and dynamic stimuli, therefore, is critical for understanding how real-world behavior and cognition emerge from brain activity, and this remains a major gap in our understanding of ASD. The purpose of the current proposal is to fill this gap, by using complex video stimuli that sample broadly from the natural world and engage multiple diverse perceptual and cognitive systems simultaneously, thus evoking activity across the entire brain at once. From this data, rich high-dimensional measures can be generated and used in combination with multivariate analytic methods ideally-suited to detect idiosyncratic and heterogeneous patterns of neural responding, which can then be related back to phenotypic variability across individuals. Our proposed studies will take place over 5 years and include 4 specific aims. In the first two aims, we will identify neural systems most affected in ASD during the presentation of a complex video stimulus, parse the heterogeneity at the neural level using data-driven approaches and relate it back to heterogeneity at the behavioral level, and explore the stimulus dimensions (social and non-social alike, both sampled broadly) that underlie these neural abnormalities. In a third aim, we will examine video-evoked functional connectivity both within and across brain networks, comparing this directly against resting-state connectivity, and examining both modes of functional connectivity across various timescales (including dynamic coupling). A final exploratory aim will assess the short-term and long-term stability of these measures, as well as their sensitivity in tracking change following experimental perturbation—important characteristics for potential biomarkers and/or predictor and outcome measures for use in intervention studies. Altogether, this work will provide new insight into brain activity and brain connectivity during conditions that more closely reflect processing demands of the natural world, help to link individual differences in brain functioning with individual differences in behavior (i.e., heterogeneity), and assess whether these neural measures may be viable candidates as biomarkers for use in future studies. This proposal addresses the Interagency...