Three hundred nineteen cervical cerclages performed in 264 pregnant women were retrospectively studied. The diagnosis of cervical incompetence was established by the obstetrical history, hysterosalpingography, ultrasound screening and vaginal examination. All cervical cerclages were applied between 14 and 17 weeks gestation according to Shirodkar's technique except 49 emergency cases between 18 and 26 weeks gestation for which other techniques were chosen. The incidence of preterm deliveries (26-37 weeks gestation) decreased from 39.7% to 14.23% (P less than 0.001) and that of full term pregnancies increased from 20.04% to 75.74% (P less than 0.001). The number of neonates weighing less than or equal to 2000 g decreased from 44.20% to 11.38% (P less than 0.001) and those weighing greater than or equal to 2500 g increased from 44.83% to 75.82% (P less than 0.001). Perinatal mortality after cerclage declined from 28.21% to 5.52% (P less than 0.001). There was no increase in congenital defects. An increased rate of breech presentation (5.32%) and cesarean section (20.38%) was noticed. The repeated cerclage in consequent pregnancies did not seem to influence the duration of gestation. In emergency cases the rate of preterm deliveries was 53.06%, of full term pregnancies 12.25%, of newborns with birthweight less than or equal to 2500 g 31.77% and of those with birthweight greater than 2500 g 19.23%. Perinatal mortality in emergency cases was 42.3%. An increase in aerobic and anaerobic pathological flora was noticed in postoperative cervical cultures.