Abstract Background: In a Phase II study, osilodrostat, a potent oral 11β-hydroxylase inhibitor, normalized mean urinary free cortisol (mUFC) in most patients with CD (Fleseriu M et al. Pituitary 2016;19:138-48). We report the efficacy and safety of osilodrostat in a large CD patient population in the first randomized withdrawal Phase III study in this rare, serious disorder (clinicaltrials.gov: NCT02180217). Methods: Phase III, multicenter, double-blind, randomized withdrawal study following a 24-week, open-label, single-arm treatment phase. Open-label osilodrostat was initiated at 2 mg bid in 137 adults with CD and mUFC (mean of three 24-h samples) >1.5xULN (ULN=50.0 µg/24h), with dose adjustments every 2 weeks (dose range 1-30 mg bid) up to week (W) 12 based on efficacy (if mUFC >ULN) and tolerability. At W26, 71 eligible patients (mUFC ≤ULN at W24 without a dose increase after W12) were randomized to continue osilodrostat (n=36) or matching placebo (n=35) for 8 weeks, followed by open-label osilodrostat until W48. Patients who remained on treatment at W26 but were not eligible for randomization continued open-label osilodrostat (n=47). Primary endpoint: patients in each randomized group with mUFC ≤ULN at the end of the randomized withdrawal phase (W34) without a dose increase after W26. For all mUFC assessments, patients who discontinued were classed as non-responders at subsequent time points. Results: At baseline, median (range) mUFC was 3.5xULN (0.3-69.6) in enrolled patients. At the end of the randomized withdrawal period (W34), significantly more patients maintained mUFC ≤ULN (without a dose increase after W26) in the osilodrostat group than in the placebo group (86% vs 29%; OR 13.7, P<0.001). Other endpoints: at W24, 53% of enrolled patients had mUFC ≤ULN without a dose increase after W12 (key secondary endpoint); at W48, 66% of enrolled patients had mUFC ≤ULN; 96% of enrolled patients had mUFC ≤ULN at least once during the study; median time to first mUFC ≤ULN was 41 days. Median (range) duration of osilodrostat exposure was 75 weeks (1-165). By W48, 24 (18%) patients had discontinued the study, 15 (11%) because of AEs. Overall, the most common AEs were nausea (42%), headache (34%) and fatigue (28%). AEs related to hypocortisolism and adrenal-hormone-precursor accumulation occurred in 51% and 42% of patients, respectively. The most common AEs reported in the osilodrostat group during the randomized withdrawal period were nausea (11% vs 0% for placebo), anemia (8% vs 9%), arthralgia (8% vs 0%) and headache (8% vs 0%). Conclusion: Osilodrostat was significantly superior to placebo at maintaining mUFC ≤ULN after randomized withdrawal and normalized mUFC in two-thirds of enrolled patients at W48, with few patients discontinuing treatment because of AEs. This randomized withdrawal study demonstrates osilodrostat to be a highly effective treatment for CD, with good tolerability.