Cervical cerclage has been shown to reduce the risk of recurrent spontaneous preterm birth in a high-risk patient population; however, the mechanism is not well understood. Transabdominal cerclage is superior to low and high vaginal cerclage in reducing early spontaneous preterm birth and fetal loss in women with previous failed vaginal cerclage. Cervical length measurements are commonly used to monitor high-risk women and may explain the mechanism of success. This study aimed to evaluate the rate of change in longitudinal cervical length after randomized placement of low transvaginal, high transvaginal, or transabdominal cerclage in women with a previous failed vaginal cerclage. This was a planned analysis of longitudinal transvaginal ultrasound cervical length measurements from patients enrolled in the Vaginal Randomised Intervention of Cerclage trial, a randomized controlled trial comparing transabdominal cerclage or high transvaginal cerclage with low transvaginal cerclage. Cervical length measurements at specific gestational ages were compared over time and between groups, using generalized estimating equations fitted using the maximum-likelihood random-effects estimator. In addition, cervical length measurements were compared in women with transabdominal cerclage placed before and during pregnancy. The diagnostic accuracy of cervical length as a predictor of spontaneous preterm birth at <32 weeks of gestation was explored. This study included 78 women who underwent longitudinal cervical length assessment (70% of the analyzed cohort) with a history of failed cerclage, of whom 25 (32%) were randomized to low transvaginal cerclage, 26 (33%) to high transvaginal cerclage, and 27 (35%) to transabdominal cerclage. Abdominal cerclage was superior to low (P=.008) and high (P=.001) vaginal cerclage at maintaining cervical length over the surveillance period (14 to 26 weeks of gestation) (+0.08 mm/week, 95% confidence interval, -0.40 to 0.22; P=.580). On average, the cervical length was 1.8 mm longer by the end of the 12-week surveillance period in women with transabdominal cerclage (+1.8 mm; 95% confidence interval, -7.89 to 4.30; P=.564). High vaginal cerclage was no better than low cervical cerclage in the prevention of cervical shortening; the cervix shortened by 13.2 mm over 12 weeks in those with low vaginal cerclage (95% confidence interval, -21.7 to -4.7; P=.002) and by 20 mm over 12 weeks in those with high vaginal cerclage (95% confidence interval, -33.1 to -7.4; P=.002). Preconception transabdominal cerclage resulted in a longer cervix than those performed during pregnancy; this difference was significant after 22 weeks of gestation (48.5 mm vs 39.6 mm; P=.039). Overall, cervical length was an excellent predictor of spontaneous preterm birth at <32 weeks of gestation (receiver operating characteristic curve, 0.92; 95% confidence interval, 0.82-1.00). In women with a previous failed cervical cerclage, in the next pregnancy, the cervical length in women treated with vaginal cerclage funneled and shortened over time, whereas there was preservation of cervical length in women who receive transabdominal cerclage. Cervical length remained longer in transabdominal procedures performed before pregnancy than in transabdominal procedures performed during pregnancy. Overall, cervical length was an excellent predictor of spontaneous preterm birth in our cohort. Our findings may explain the mechanism of benefit for transabdominal cerclage, with its high placement better maintaining the structural integrity of the cervix at the level of the internal os.