PROJECT SUMMARY/ABSTRACT Autism spectrum disorder (ASD) is a severe and heterogeneous neurodevelopmental disorder characterized by core deficits in social and communication functions. In trying to delineate the underlying neural mechanisms of ASD, neuroimaging studies have focused on networks subserving social-emotional perception and social cognitive processes such as theory of mind (ToM). The role of social-emotional regulation (ER), a central component of effective and coherent social-emotional functioning with treatment implications, in ASD is understudied, albeit recent evidence points to abnormal ER as a potential mechanism for ASD core and non- core (e.g. anxiety) symptoms. Although understanding the neural networks underlying social-emotional perception, cognition and regulation has advanced, their temporal and structural architecture is not fully characterized in the typically developed brain and in ASD. This can be partially attributed to current methods that (1) ignore the dimensionality of ASD related symptoms, and (2) measure functional connectivity (FC; a measure characterizing neural networks) as a static process, ignoring the temporal dynamic interactions within and between networks. The current proposal focuses on the dynamic FC of the neural networks subserving social-emotion processing and regulation of happy and sad simulated social interactions as measured by functional MRI (fMRI) during an ecologically valid task. Two-hundred adults ages 18 to 40, 50 with high- functioning ASD and 150 mixed, non-ASD controls, will undergo a comprehensive assessment of social processing, ER as well as internalizing symptoms (e.g. depression and anxiety). During two MRI sessions they will view short video clips of actors talking about happy, sad or neutral situations while looking directly at the camera, simulating real-life social interactions. While the first session will entail passive viewing, during the second session participants will be asked to apply explicit ER reappraisal strategies they will be trained on. We will use multivariate validated approach that includes independent component analysis (ICA) and dynamic functional network connectivity (dFNC) to assess the temporal dynamic FC within and between social- emotional networks, and domain transition analysis (DTA) to assess the influence of explicit ER on these dFNC patterns (i.e. FC states). Since social processes and functioning, including ER, autistic traits and other related symptoms (e.g. anxiety) are dimensional constructs, we will take a dimensional (vs. categorical) analysis approach to test their predictive values on dynamic FC. If successful, this study will shed light on the neurobiological underpinnings of social-emotional processing and regulation across ASD and healthy individuals, with implications to development of new treatment strategies in this and other psychiatric populations.