Summary: The aim of this study was to explore the hypothesis that serial transvaginal ultrasonography identifies early evidence of suture failure and that repeat cerclage delays delivery. We undertook a review of our policy of transvaginal ultrasonographic cervical surveillance after McDonald cerclage and of repeat suture insertion if persistent cervical effacement developed. Data from 26 pregnancies in 26 women are analyzed. The women had had a total of 57 mid‐trimester miscarriages with a median of 2 (1–6) mid‐trimester losses per woman. Twelve (46%) of the 26 women developed cervical changes at scan and underwent repeat cerclage. All 14 women who had a single suture inserted progressed to live births but 1 of the 13 women who had repeat cerclage had a mid‐trimester miscarriage (p>0.05). The median gestation at delivery for the women who had repeat cerclage was 35 (22–39) weeks compared with 38 (36–40) weeks for those who had a single suture (p>0.05). The median interval from the detection of cervical changes at scan to delivery was 13 (4–19) weeks. Serial transvaginal ultrasonography after cervical cerclage identifies a group of women who are more likely to deliver preterm, and provides an opportunity for intervention (repeat cerclage) which appears to delay delivery by an average of 7 weeks.