PURPOSE: We aimed to assess the efficacy of biofeedback therapy for patients with fecal incontinence (FI) according to the etiology. METHODS: Twenty-nine patients with fecal incontinence were treated with biofeedback therapy using a EMG-based system. The efficacy was assessed by using changes in the FI score (Cleveland Clinic, Florida: 0~20) and satisfaction based on a subjective evaluation score from 0 to 100. The median follow up duration was 12 (3~25) months. RESULTS: Ten patients had idiopathic fecal incontinence. Fourteen patients had fecal incontinence due to a sphincter saving operation for rectal cancer. Four cases had spinal cord injury and one patient had a major external sphincter tearing due to trauma. The mean age was 52 (16~78) years. The median number of biofeedback sessions was 10 (3~15) overall. The mean efficacy was 42.8%, and the mean satisfaction score was 56.6. Improvements in the FI score and in the patients' satisfaction varied according to the etiology, 69.5% and 71.5 in the idiopathic group, 28.5% and 49.3 in the postoperative group, and 35% and 24 in the spinal cord injury group. In the idiopathic group, 50% of the patients showed an improvement in the FI score of more than 75%, and 90% of the patients showed an improvement of more than 50%. The number of liquid incontinence episodes was improved 78.3% later in the biofeedback group, and this result was much better than in the postoperative incontinence group (31.8%, p=0.03). CONCLUSIONS: The success rate of the biofeedback therapy for fecal incontinence is acceptable. Subjective satisfaction is relatively higher than the improvement in the ecal incontinence score. Idiopathic fecal incontinence may be the best indication for biofeedback therapy.