The objective of this study was to compare mesh erosion after transvaginal repair of cystocele using Gynemesh or Gynemesh-Soft mesh. We retrospectively analyzed 138 consecutive cases of transvaginal repair of cystocele using synthetic mesh. The study endpoint was the pathological evidence of vaginal erosion. Multiple logistic regression was used to determine independent predictors of vaginal erosion. One hundred and thirty eight women (ages 30-83 years) with cystocele between October 1999 and October 2004, from a French University Hospital, participated in this study. Cystocele repair was performed in all patients according to the technique of tension-free polypropylene mesh. The median follow-up was 32.1 months (range 7.5-59.9) in the Gynemesh group and 7.1 months (range 1-21.9) in the Gynemesh-Soft group. Vaginal erosion was reported in 27 (20%) of the patients. Anatomically, the success rate was 95% (131/138). There was no statistically significant difference between the Gynemesh and the Gynemesh-Soft meshes [the rate of vaginal erosion of the mesh was 16% (15/89) vs 24% (12/49), respectively, p=0.39]. Univariate analysis only identified age class as factor significantly associated with the probability of vaginal erosion. Multivariate analysis revealed that age class is an independent predictive factor of vaginal erosion (age > 70 years, odds ratio (OR) 3.6, 95% confidence interval (CI) 1.3-9.7, p=0.010). Furthermore cystocele stage > 2 (Baden and Walker classification) is a protective factor against vaginal erosion (OR 0.3, 95% CI 0.1-0.8, p=0.016). Thirteen symptomatic patients (13/27, 48%) necessitated a partial excision of the mesh, associated with a vaginal mucosal closure. Two patients (2/27, 7%) underwent a complete excision of the mesh. The incidence of de novo dyspareunia was 9% in patients with vaginal erosion and 11% in patient without mesh erosion (p=0.85). There was no occurrence of bladder or urethral erosion and no vaginal or pelvic infection. Isolated vaginal erosion of the mesh did not prove to be problematic. Gynemesh-Soft mesh does not decrease the incidence of vaginal erosion. Age > 70 years is an independent predictive factor of vaginal erosion. We recommend that mesh placement by vaginal route should be avoided by women with moderate cystocele. Where possible, total hysterectomy and vertical incision should also be avoided. Management of vaginal erosion is simple and is associated with a low rate of morbidity. However, patients should be informed that vaginal erosion of the mesh can occur. A multivariate analysis reveals that the incidence of vaginal erosion is not significantly different between Gynemesh and Gynemesh-Soft meshes. Other factors of erosion are analyzed.