Although surgical therapy has been shown to be an effective treatment of anal incontinence, few properly controlled randomized studies have confirmed its efficacy or compared it with biofeedback or other less invasive forms of treatment. Overlapping sphincteroplasty, the most common procedure, seems to confer substantial benefits on patients with sphincter disruptions. However, recent data suggest that results following sphincteroplasty deteriorate with time. There is also disagreement about whether pudendal nerve conduction studies can be used to predict outcome after surgical repair. Salvage options for patients with refractory fecal incontinence include passive or electrically stimulated muscle transfer procedures, implantation of an inflatable artificial anal sphincter, and sacral nerve stimulation. Stimulated graciloplasty is the most commonly used muscle transfer procedure; good to excellent results are reported from a small number of high-volume centers, but multicenter trials with less experienced surgeons have shown a high morbidity rate associated with the procedure. The artificial anal sphincter provides good restoration of continence for most patients who retain the device, but a significant explantation rate due to infection or local complications remains problematic. Sacral nerve stimulation has shown promising early results with minimal associated morbidity. There is a critical need for controlled long-term studies that use objective data collection methods, standardized outcome measures, and validated quality-of-life assessment instruments.