This review presents reported cure and improvement rates of stress urinary incontinence in women obtained by different treatment modalities. Apart from the urodynamic findings, histological and histochemical changes of the pelvic floor may be clinically relevant to treatment in the future. Long-term cure and improvement rates achieved by non-surgical treatment (physiotherapy, biofeedback, bladder training, electrostimulation) are commented on. These rates range from 40-60% for physiotherapy and electrostimulation but are considerably less after biofeedback and bladder training. Pharmacotherapy is unlikely to offer more than a placebo effect. Studies of a single surgical procedure usually report high cure rates. In making the appropriate choice of operation the best guidelines are the cure rates from comparative or prospective randomized reports. From such studies an abdominal retropubic suspension operation (cure rates after five years 57-78 %) is more likely to help the patient than an anterior colporrhaphy (cure rates 31-70 %) or a transvaginal needle bladder neck suspension (cure rates 39-61 %). In selected patients sling procedures or the use of artificial sphincters may produce excellent results (70-80 %). To estimate the results of different treatments urine loss should be assessed objectively and physical restrictions and hygienic and social implications taken into account. A method of pre- and post-treatment "performance scores" should be developed.