Study design: Retrospective analysis. Objectives: To evaluate long-term efficacy and complications of AMS 800 (American Medical Systems s ) artificial urinary sphincter (AUS) in treatment of urodynamic stress incontinence (USI) in male patients with spinal cord lesion (SCL). Setting: London Spinal Injuries Unit, Stanmore, UK and Institute of Urology and Nephrology, London, UK. Patients and methods: A retrospective analysis identified nine males with SCL (five thoracic, three lumbar, one cervical) with USI, who underwent AMS 800 AUS implantation by a single surgeon at a specialist spinal injuries unit. The mean age was 38.2 years (range 27-47 years), with the mean time since injury of 13.8 years (range 6-26 years). In all, seven were complete and two were incomplete SCL (traumatic spinal injury eight, tranverse myelitis one). All implants were inserted with the urethral cuff around the bulbar urethra with a 61-70 cm water pressure reservoir in the retropubic space. Implant activation was carried out at 6 weeks postoperatively. All patients were regularly followed up in outpatient clinics at the interval of 3 months, 6 months and yearly thereafter. An ultrasound examination for the upper tracts and a videocystometrogram (VCMG) was carried out at 3 months postsurgery and then yearly. Results: The follow-up ranged from 3 to 133 months (mean 70.2 months). All implants were activated successfully with no intra-or immediate-postoperative complications. At activation all patients reported total urinary continence with seven out of nine implants (77%) currently working well. Two patients reported significant recurrent incontinence at 3 month follow-up, one of whom underwent a removal of the entire implant at the end of 3 months and the other was continent after a pump and cuff revision. The implant removed at 3 months was due to erosion and infection while the second was removed at 24 months due to secondary implant infection. Three out of seven (43%) successful implants required one revision each. One patient continues to report minimal leakage only during transfers with no leak demonstrated on postoperative VCMG. One patient with indwelling urethral catheter, two with suprapubic catheter and one voiding on urge have changed their bladder management to intermittent catheterisations postoperatively. All removals and revision procedures were carried out in the first 53 months of follow-up and four out of seven implants (57%) required no revisions. Conclusion: On a long term, AMS 800 is a viable option to treat USI in men with SCL. Mechanical revisions are frequent but effective. Strict patient selection, optimum preoperative bladder management and regular follow-up ensure low complication and high efficacy rates in the long term.