Cushing's disease is a rare debilitating endocrine disorder for which few prospective interventional studies have been done. We report results of the first phase 3 trial assessing long-acting intramuscular pasireotide in patients with Cushing's disease. In this phase 3 clinical trial we recruited patients aged 18 years or older with persistent, recurrent, or de-novo (non-surgical candidates) Cushing's disease who had a mean urinary free cortisol (mUFC) concentration (from three 24 h samples) of 1·5-5·0 times the upper limit of normal (ULN), a normal or greater than normal morning plasma adrenocorticotropic hormone concentration, and a pituitary source of Cushing's syndrome, from 57 sites across 19 countries. Exclusion criteria included previous pasireotide treatment, mitotane therapy within 6 months, and pituitary irradiation within 10 years. We randomly allocated patients 1:1 (block size of four) using an interactive-response-technology system to intramuscular pasireotide 10 mg or 30 mg every 4 weeks for 12 months (in the core phase). We stratified randomisation by screening mUFC concentration (1·5 to <2·0 × ULN and 2·0-5·0 × ULN). The dose could be uptitrated (from 10 mg to 30 mg or from 30 mg to 40 mg) at month 4 if the mUFC concentration was greater than 1·5 × ULN, and at month 7, month 9, or month 12 if the mUFC concentration was greater than 1·0 × ULN. Investigators, patients, site personnel, and those assessing outcomes were masked to dose group allocation. The primary endpoint was the proportion of patients in each group with an mUFC concentration of less than or equal to the ULN at month 7. Efficacy analyses were based on intention to treat. This trial is registered with ClinicalTrials.gov, number NCT01374906. Between Dec 28, 2011, and Dec 9, 2014, we randomly allocated 150 patients to receive pasireotide 10 mg (74 [49%] patients) or 30 mg (76 [51%] patients). The primary efficacy endpoint was met by 31 (41·9% [95% CI 30·5-53·9]) of 74 patients in the 10 mg group and 31 (40·8% [29·7-52·7]) of 76 in the 30 mg group. The most common adverse events were hyperglycaemia (36 [49%] in the 10 mg group and 36 [47%] in the 30 mg group), diarrhoea (26 [35%] and 33 [43%]), cholelithiasis (15 [20%] and 34 [45%]), diabetes mellitus (14 [19%] and 18 [24%]), and nausea (15 [20%] and 16 [21%]). Serious adverse events suspected to be study drug related were reported in eight (11%) patients in the 10 mg group and four (5%) in the 30 mg group. Two (3%) patients in the 30 mg group died during the study (pulmonary artery thrombosis and cardiorespiratory failure); neither death was judged to be related to the study drug. Long-acting pasireotide normalised mUFC concentration in about 40% of patients with Cushing's disease at month 7 and had a similar safety profile to that of twice-daily subcutaneous pasireotide. Long-acting pasireotide is an efficacious treatment option for some patients with Cushing's disease who have persistent or recurrent disease after initial surgery or are not surgical candidates, and provides a convenient monthly administration schedule. Novartis Pharma AG.