A prospective, single-centre study to assess the outcome of incontinence surgery in the first 120 consecutive patients who had tension-free vaginal tape (TVT) by a single surgeon. All patients were initially seen at 3 months postsurgery, with a cough provocation test, measurement of residual urine volume and a satisfaction survey. At a mean of 26 months (6-42 months) after surgery, a validated telephone interview was performed. The operation was performed in accordance with the original technique described by Ulmsten et al. [Int Urogynecol J Pelvic Floor Dysfunct 1996; 7: 81-5]. A total of 87 of 120 patients completed the study with the others either not complying or having died. Sixty-three (72.4%) patients were completely dry on cough provocation test. Of these, four (4.5%) had a slow stream and 10 (11.4%) suffered persistent urgency. The remaining 24 patients had varying degrees of leakage (operative failure). Sixteen (18.3%) patients subjectively considered the procedure to have failed at 3 months follow-up, either because leakage occurred once or more a day, and/or the persistence of the preoperative frequency/urgency syndrome. Of these 16 TVT failures, two had previous pelvic radiotherapy, two had double incontinence and eight had TVT for recurrent incontinence. Among the failures, 81.3% had mixed incontinence with predominant urge and nocturia three times per twenty four hours. Our study highlights the need for selection when performing TVT. We recommend that TVT be performed for those who have simple stress incontinence failing conservative measures (pelvic floor exercises and physiotherapy), with no history of incontinence surgery, pelvic radiotherapy, faecal or mixed incontinence.