PROJECT SUMMARY Eating disorders (ED) are serious mental illnesses with few evidence-based treatments and little mechanistic understanding of the neural circuit impairments that underlie both illness and recovery. This project contributes to developing a more dimensional understanding of EDs, by comparing the neural circuits involved in both self- perception and social reward in a previously unstudied clinical cohort, atypical anorexia nervosa (AAN) to a psychiatrically-healthy cohort that has lost similar amounts of body weight (WLC). Typical anorexia nervosa (AN), for the purpose of research and clinical care, is defined by a malnourished body state (BMI < 18.5) with persistent behaviors that interfere with weight restoration, and often includes preoccupations with body image. Atypical anorexia nervosa (AAN) was defined in DSM-5 as meeting all criteria for AN without being underweight; these patients have typically experienced substantial weight loss due to severe restrictive eating behaviors as well as disruptions of their psychosocial functioning due to the severity of their ED cognitions. Clinically, AAN is associated with similar to higher levels of psychological distress and distortions about both eating and body- image as typical AN. This exploratory pilot project will evaluate whether similar neurobehavioral problems in self- perception and social reward are observed in the brain in AAN, extending an existing R01 that is currently examining these circuits in both AN and bulimia nervosa (BN). In addition, the psychiatrically-healthy weight loss cohort (WLC) will have experienced significant weight loss (10-20% of highest body weight) but will not endorse pathological ED cognitive preoccupations; this cohort, in comparison to AAN, allows separation of the impact of weight loss on brain function from the neuropsychological processes that mediate body preoccupations. In sum, this exploratory pilot study evaluates neurobiological circuitry in an understudied diagnosis, AAN, and provides the necessary data to fill current gaps in our knowledge about the effects of weight and ED cognitions on neural function as well as clinical symptoms in EDs. These data are a step towards a more mechanistic understanding of the pathophysiology in EDs.