The purpose of this study was to identify the determinants of choice of surgical procedure (anterior colporrhaphy, colposuspension, or needle suspension) to treat stress incontinence in women. We used multilevel modeling of data on 271 patients in 18 hospitals in England in 1993–94. Patient-related factors included sociodemographic details, anatomical diagnosis, symptom severity, symptom impact, previous treatment, parity, comorbidity, and general health status. Surgeon-related factors were specialty, grade, and annual volume of procedures undertaken. Hospital teaching status was considered. Some patient-related factors were associated with choice of procedure: women with a concomitant genital prolapse, with a history of high parity, and with no previous nonsurgical treatment were more likely to undergo an anterior colporrhaphy than a colposuspension or needle suspension (although this finding could be confounded by surgical specialty). In addition, women were more likely to be treated by colposuspension if their surgeon specialized in incontinence surgery (measured by annual volume of cases). Finally, being treated by needle suspension depended on there being a consultant surgeon familiar with the procedure at the hospital attended. While choice of surgical procedure depends partly on the patient's anatomical diagnosis, it is also dependent on the specialty of the surgeon whom she consults and the hospital that she attends. This variability, in turn, could have implications for the patient (as the relative effectiveness of the different procedures is unknown) and for the purchasers of care (as the relative cost-effectiveness of procedures is also unknown).