OBJECTIVE: Gastrointestinal ultrasound (GIUS) accurately assesses inflammation and is responsive to changes in inflammatory bowel disease. This study aimed to determine the prognostic utility of sonographic response in the first 14 weeks of a newly-instituted therapy with therapeutic response at 46 weeks and to compare its performance with standard clinical assessment tools. METHODS: Patients with sonographic evidence of inflammation were assessed by GIUS, clinical activity, serum C-reactive protein and faecal calprotectin again 2, 6 and 14 weeks after commencing a new biologic or thiopurine. Treatment failure was defined as undergoing surgery, hospitalisation, escalation of dosage or introduction of new medication over 46-weeks' follow-up. Sonographic response was defined as a decrease in bowel wall thickness and improved vascularity. RESULTS: In 31 patients (median age 49 years, 74% Crohn's disease), sonographic response at 14 weeks [OR 19.3, 95% confidence interval (CI), 3.23-101.10; P = 0.0054] and faecal calprotectin (P = 0.018), but no clinical disease activity or C-reactive protein, were predictive of subsequent treatment response. Sonographic response alone was predictive at week 6 (P = 0.016), but not week 2. 16% reduction in bowel wall thickness at 6 weeks (area-under-the-receiver-operator-curve=0.86; P = 0.002; sensitivity 72%, specificity 90%), with similar performance for 10% at 14 weeks, was associated with treatment response. CONCLUSION: Sonographic response as early as 6 weeks after initiation of a new therapy may accurately predict treatment outcomes over 46 weeks and is superior to other markers used to monitor disease activity.