OBJECTIVE We evaluate the long-term outcome of the Gittes procedure for urinary stress incontinence. METHODS A total of 87 women with proved genuine stress incontinence were treated with the Gittes procedure. The same urologist performed 95 consecutive operations during an 8-year period. Patients were evaluated by a postal questionnaire. RESULTS Of the patients 52 (60%) (55 operations) responded to the questionnaire. Mean and median followup were 53 and 46 months, respectively (range 24 to 103). Twelve patients (23.1%) reported complete absence of postoperative urinary incontinence and were considered cured, 14 (26.9%) were significantly improved and a total of 30 (57.7%) benefited from the operation. The short-term results were initially encouraging but by 2 years only 20 patients were completely continent (38.5% cured). Of the 40 patients who were not cured 32 (80.0%) experienced incontinence within 2 years postoperatively. There were 26 who had complained of frequency and/or urgency preoperatively. There was a statistically significant subjective failure rate in this group (p = 0.007). CONCLUSIONS The Gittes procedure is simple and has minimal complications. Although it provides continence in the early weeks and months following surgery, the long-term cure rate is disappointing, with most failures occurring within 2 years of surgery. Preoperative irritative symptoms, even when multichannel cystometry did not reveal instability, were associated with a poor subjective outcome. Our results suggest that the Gittes procedure is not satisfactory for the management of genuine stress incontinence in women.