Avoidant restrictive food intake disorder (ARFID) is a new psychiatric disorder in the Diagnostic and Statistical Manual 5 (DSM-5). ARFID has an estimated prevalence of 7.2 to 17.4 percent thus making it a significant mental health concern. ARFID is characterized by a range of dysfunctional eating behaviors including a lack of interest in eating, sensory related eating concerns (such as taste, color or texture) and a fear of adverse consequences of eating (i.e., fear of choking or vomiting). There is no evidence-based treatment for ARFID. Preliminary data from a feasibility study comparing FBT-ARFID to Usual Care (UC) provide evidence that manualized FBT adapted for patients with ARFID is feasible and effective. Recruitment and randomization averaged 1.87 participants per month over a 15 month period with an overall attrition rate of 21%, comparable to rates in fully powered studies of FBT-AN. The feasibility study also identified an efficacy signal on the difference between groups on the primary outcome (change in percent Estimated Body Weight (%EBW)) of a large effect size (ES) favoring FBT-ARFID Studies suggest that improvements in parental self-efficacy related to changing feeding and eating behaviors early in treatment is a likely mechanism of FBT for other eating disorders in youth. Our feasibility study showed a striking difference between conditions in parental self-efficacy favoring FBT-ARFID compared to UC. In addition to this promising evidence of target engagement In addition, target validation was demonstrated by the change in parental self-efficacy being significantly correlated with improvements in % EBW. Aim 1: To conduct an RCT involving children and adolescents between the ages of 6 and 12 years of age with DSM 5 ARFID and weight below 88% of EBW comparing FBT-ARFID with medical management to manualized Non-Specific Treatment UC with medical management. Treatments will be matched for time and therapist attention. We hypothesize that participants randomized to FBT-ARFID will have significantly greater change in %EBW at EOT. Aim 2: To examine early change in parental self-efficacy as a mediator of treatment effect (FBT-ARFID vs. UC on outcome). We hypothesize that positive changes due to FBT-ARFID in parental self-efficacy related to feeding behaviors using the Parents vs ARFID Scale (PvsARFID) will be associated with positive changes in %EBW at EOT. Secondarily, we will explore whether objective changes in parental re-feeding behavior is a possible mechanism of FBT-ARFID using a mediator analysis. Aim 3: To explore moderators of treatment outcome. To conduct an adequately powered study, 100 children (ages 6-12 years) will be randomized to manualized FBT-ARFID plus medical management (n=50) or manualized Non- Specific Treatment plus medical management (n=50). Assessments (blinded to treatment condition) of primary and secondary outcomes will be conducted at baseline, 1 month, 2 months, and 4 months (EOT).