Anorexia nervosa (AN) is a serious mental illness with significant morbidity and mortality rates, and relapse rates ranging from 40-80% after acute treatment. A central behavioral disturbance in AN is the restriction of food intake, which is promoted by entrenched thoughts, feelings, and routines. While many approaches to reducing relapse after hospital care have been tried, there is little information about which treatment elements confer benefit. Relapse Prevention and Changing Habits (REACH+) targets the habitual control of maladaptive behavior to support patients with AN in the 6 months after acute treatment, a time of high vulnerability to relapse. REACH+ builds on preliminary data that indicate that behavioral interventions that target habits in AN can lead to changes in entrenched behaviors in the acute treatment setting. This approach has not yet been developed for the relapse prevention setting. We will test whether treatment components can interrupt the habitual control of maladaptive behavior and, by targeting this illness mechanism, decrease relapse and improve outcomes. We will use a Multiphase Optimization STrategy (MOST) framework an ideal means for improving treatment efficacy by simultaneously and efficiently evaluating multiple possible intervention components, to determine which contribute to positive treatment outcomes and should therefore be included in a robust treatment package for relapse prevention. Each component of REACH+ addresses a question that is critical to answer in order to identify and optimize a relapse prevention treatment package that balances efficacy and burden. We will test competing versions of 5 components that together target habits: 1) Behavioral, 2) Cognitive, 3) Motivation, 4) Food Monitoring, and 5) Skill Consolidation. First, we will conduct the final stages of the Preparation Phase (including Focus Groups) to maximize feasibility and acceptability of the intervention components. Second, an Optimization Phase experiment with a finalized treatment manual (including an online platform) will test each component's contribution to weight maintenance after acute treatment. We will provide 6 months of REACH+ for 60 individuals with AN, weight-restored via our Eating Disorders Program. Sessions, conducted via telehealth, will be augmented with psychoeducation, homework assignments, and monitoring through an online platform. Participants will be randomly assigned to different versions of each treatment component. Optimization criteria focus on the contribution of the component to weight maintenance (main effect), and will additionally consider effect sizes, treatment adherence, and end-of- treatment status. This will ensure that the finalized treatment package includes the most effective and feasible components for definitive testing in a future randomized controlled trial. This systematic approach to treatment development is critical for achieving the best outcomes most quickly and efficiently.