ABSTRACT Avoidant/restrictive food intake disorder (ARFID) affects 3% of children and adolescents and results in nutritional deficiencies, supplement dependence, and psychosocial impairment. ARFID follows a chronic course, and has no evidence-based treatment. The hallmark feature of ARFID is food avoidance, which may be maintained by extreme levels of food neophobia and/or hyperactivation of fear circuitry in response to food cues. Our team has developed a manualized cognitive-behavioral therapy (CBT-AR) that directly targets both food neophobia (cognitive mechanism) and fear circuitry (neural mechanism) to reduce food avoidance (clinical outcome). In line with NIMH’s experimental therapeutics approach, in this exploratory/developmental phased R61/R33, we will leverage a multidisciplinary team of experts in the treatment of ARFID, neural mechanisms of food motivation, and statistical analysis of clinical trials to conduct a mechanistic randomized controlled trial of CBT-AR. First, to establish target engagement, we will randomize 50 youth (ages 10-18yo) with ARFID in a 1:1 ratio to 15 weekly sessions (via telehealth) of CBT-AR vs. nutrition counseling to establish target engagement. We chose nutrition counseling as our active control because it does not include the crucial CBT-AR intervention of exposure, and is therefore unlikely to engage our target mechanisms. We hypothesize that, compared to nutrition counseling, patients randomized to CBT-AR will show significantly greater decreases in food neophobia (cognitive mechanism) and fear circuitry (neural mechanism) in response to food cues during a standardized fMRI food cue paradigm (primary ROI: anterior cingulate cortex [ACC]; secondary ROIs: amygdala, orbitofrontal cortex [OFC]). We will also examine weekly change in food neophobia (cognitive mechanism) to identify the number of sessions at which further benefit ceases, and use this optimized dose of CBT-AR for the R33. We will move on to the R33 if we are able to demonstrate a reduction of at least d=.40 in the CBT-AR group from pre- to post-treatment AND a post-treatment between- group difference of at least d=.40 in CBT-AR vs. nutrition counseling in either food neophobia (cognitive mechanism) OR fear circuitry (Go/No-Go ROI: ACC; neural mechanism). Next, we will randomize 70 youth (ages 10-18yo) with ARFID in a 1:1 fashion to the optimized dose of CBT-AR or the same number of sessions of nutrition counseling to replicate target engagement and establish target validation. We hypothesize that, compared to nutrition counseling, patients randomized to CBT-AR will exhibit significantly greater reductions in food avoidance, and that these reductions in food avoidance will be mediated by reductions in food neophobia (cognitive mechanism) and fear circuitry (neural mechanism) activation. If successful, the proposed intervention (CBT-AR) could fill an important unmet need for those living with ARFID and pave the way for larger-scale efficacy and e...