Voiding colpo-cysto-urethrography was performed in 52 consecutive female patients with genuine urinary stress incontinence before treatment and in 50 of the patients after treatment. The patients were randomized to either pelvic floor training or surgery. Surgery included a colposuspension operation in patients with an anterior suspension defect and a vaginal repair in patients with a posterior suspension defect. All pre- and posttreatment examinations were reevaluated blindly by one observer 4 to 6 years later. The pretreatment radiologic reevaluation was in agreement with the original classification in 79 per cent and not in agreement in 21 per cent. Pelvic floor training did not change the degree of suspension defect systematically. The effect of squeezing was significantly improved following pelvic floor training. A colposuspension gave rise to a typical radiologic configuration of the bladder and urethra. A vaginal repair was not detectable radiologically and it did not correct a posterior descent. The degree of descent and the degree of incontinence were not correlated and it was not possible radiologically to distinguish treatment failures from treatment success.
