BACKGROUND & AIMS: Irritable bowel syndrome (IBS) is a functional gastrointestinal disorder associated with intestinal dysbiosis. Given the reported promising results of open-label fecal microbiota transplantation (FMT) therapy in patients with predominant abdominal bloating, we studied efficacy of this treatment in a randomized, placebo-controlled trial. METHODS: Patients with refractory IBS, defined as failure of ≥3 conventional therapies, were randomly assigned to single-dose nasojejunal administration of donor stools (n = 43) or autologous stools (n = 19) in a double-blind study, performed from December 2015 through October 2017, and were followed up for 1 year. IBS-related symptoms were assessed by using a daily symptom diary to determine general abdominal discomfort, abdominal bloating, abdominal pain, and flatulence on a scale of 1-6. Number of daily bowel movements, consistency of the stools, and abdominal circumference were also recorded. Patients completed the IBS-specific quality of life questionnaire. Primary endpoints were improvement of IBS symptoms and bloating at 12 weeks (response). Secondary endpoints were changes in IBS symptom scores and quality of life. Stool samples were collected for microbiota amplicon sequencing. Open-label retransplantation was offered after the trial. RESULTS: At week 12, 56% of patients given donor stool reported improvement in both primary endpoints compared with 26% of patients given placebo (P = .03). Patients given donor stool had significant improvements in level of discomfort (mean reduction, 19%; median score before FMT, 3.98; range, 2.13-6.00; median score after FMT, 3.1; range, 951.29-5.90), stool frequency (mean reduction, 13%; median score before FMT, 2.10; range, 0.57-14.29; median score after FMT 1.7; range, 0.71-4.29), urgency (mean reduction, 38%; median score before FMT, 0.61; range, 0.00-1.00; median score after FMT, 0.37; range, 0.00-1.00), abdominal pain (mean reduction, 26%; median score before FMT, 3.88; range, 1.57-5.17; median score after FMT, 2.80; range, 1.14-4.94), flatulence (mean reduction, 10%; median score before FMT, 3.42; range, 0.71-6.00; median score after FMT, 3.07; range, 0.79-4.23), and quality of life (mean increase, 16%; median score before FMT 32.6; range, 11-119; median score after FMT, 43.1; range, 32.25-99). A significantly higher proportion of women given donor stool (69%) had a response than men (29%) (P = .01). Fecal samples from responders had higher diversity of microbiomes before administration of donor material than fecal samples from nonresponders (P = .04) and distinct baseline composition (P = .04), but no specific marker taxa were associated with response. After single FMT, 21% of patients given donor stool reported effects that lasted for longer than 1 year compared with 5% of patients given placebo stool. A second FMT reduced symptoms in 67% of patients with an initial response to donor stool but not in patients with a prior nonresponse. CONCLUSIONS: In a randomized trial of patients with treatment-refractory IBS with predominant bloating, FMT relieved symptoms compared with placebo (autologous transplant), although the effects decreased over 1 year. A second FMT restored the response patients with a prior response. Response was associated with composition of the fecal microbiomes before FMT; this might be used to as a biomarker to select patients for this treatment. ClinicalTrials.gov, Number: NCT02299973.