BACKGROUND: Anal sphincter surgery for faecal incontinence is associated with significant morbidity and a variable outcome. Sacral nerve stimulation may provide a good functional outcome with minimal morbidity. This paper reports the experience in a single centre over 5 years. METHODS: Fifteen consecutive patients (14 women), median age 60 (range 37-71) years, underwent temporary, and subsequent permanent, stimulation. All had incontinence to solid or liquid stool; the aetiology was obstetric injury (seven patients), scleroderma (four), idiopathic (two), fistula surgery (one) and repaired rectal prolapse (one). Median duration of symptoms was 6 (range 2-15) years. Clinical evaluation, endoanal ultrasonography, bowel diary, quality of life questionnaire (Short Form (SF) 36) and anorectal physiological testing were performed before and after stimulation. RESULTS: Continence had improved in all patients at median follow-up of 24 (range 3-60) months. Eleven patients were fully continent. Episodes of faecal incontinence decreased from median (range) 11 (2-30) per week before stimulation to 0 (0-4) per week after permanent stimulation (P < 0.001). Urgency improved in all patients (median (range) ability to defer less than 1 (0-1) versus 8 (1-15) min; P = 0.01). 'Social function' and 'role-physical' subscales of the SF36 improved significantly. Mean resting pressure (35 versus 49 cmH2O with temporary stimulation; P < 0.05) and squeeze pressure increment (43 versus 69 cmH2O with permanent stimulation; P < 0.01) increased. Rectal sensitivity to initial distension changed (mean 47 versus 34 ml air; P < 0.05). There were no major complications. CONCLUSION: Sacral nerve stimulation is a safe and effective treatment for faecal incontinence when conventional treatment has failed. There is minimal morbidity. The benefit is maintained in the medium term.