Objective: Test whether a dissonance-based transdiagnostic eating disorder treatment reduces valuation of the thin beauty ideal and high-calorie binge foods, the intervention targets, and eating pathology relative to waitlist controls. Method: Women with Diagnostic and Statistical Manual of Mental Disorders–5 eating disorders (N = 100) were randomized to an 8-week group-implemented Body Project Treatment (BPT) redesigned to encourage rapid symptom reduction or a waitlist control condition, completing functional MRI paradigms assessing neural response to thin models and binge foods, questionnaires, and diagnostic interviews at pretest and posttest. Results: Compared to controls, BPT participants showed greater reductions in responsivity of regions involved in reward valuation (ventromedial prefrontal cortex, dorsolateral prefrontal cortex, caudate) to thin models but not binge foods, pursuit of the thin ideal (d = .72), palatability ratings of binge foods (d = .78), and greater increases in attractiveness ratings of average-weight models (d = .44), the intervention targets. BPT participants also showed significantly greater reductions in body dissatisfaction (d = .83), negative affect (d = .76), and eating disorder symptoms (d = .59), and marginally greater abstinence from binge eating and compensatory behaviors (39% vs. 21%) than controls. Conclusions: Results provide novel evidence that BPT affected the hypothesized intervention targets and reduced variables that are putatively secondary to pursuit of the thin ideal, including body dissatisfaction, negative affect, and eating disorder symptoms. Symptom reductions were smaller than in past trials, suggesting that it may be optimal to reduce valuation of the thin ideal before asking participants to reduce disordered eating behaviors that are used to pursue this ideal.