Pad test, cystometry and analysis of micturition were performed in 36 women with stress incontinence before and a year after operation with either retropubic urethrocystopexy (n = 22) or pubococcygeal repair (n = 14). There was no difference in the subjective cure rate between the two groups of women (77% and 79%, respectively). The pad test 1 year after operation showed that 59% of the women in the urethrocystopexy group and 43% of the women in the pubococcygeal repair group had stopped leaking urine. The bladder volume had increased in both groups and the intravesical pressure of the bladder filled to the maximum had increased in the pubococcygeal repair group. The functional length of the urethra, intravesical pressure at maximal urine flow, maximal urine flow rate and urethral conductance were not affected by either operation. Pad test was a more accurate test for an objective evaluation of urine leakage before operation than were the urodynamic investigation or continence tests.