OBJECTIVE: To evaluate the effect of cerclage, with and without cervical occlusion. DESIGN: Multicentre, stratified, randomised controlled trial. SETTING: Hospital-based multicentre study with 18 tertiary centres from nine countries. POPULATION: Women with a history of cervical insufficiency (prophylactic trial) and women with a short cervix (therapeutic trial) were recruited from August 2006 to August 2011. METHODS: A centralised telephone randomisation service with a computer system was used to randomise women to cervical cerclage with or without cervical occlusion. Only the analyst performing the interim analyses was blinded. MAIN OUTCOME MEASURES: The take-home baby rate (number of infants discharged alive from the hospital), gestational age at delivery, and the number of days in the neonatal intensive care unit (NICU). RESULTS: Women (n = 309) were stratified into the prophylactic trial (n = 213) or the therapeutic trial (n = 96). The trial stopped early due to slow recruitment and an interim analysis showing no benefit of occlusion. Final analysis comprised 197 women in the prophylactic trial and 87 women in the therapeutic trial. No added effect of cervical occlusion was found in terms of the take-home baby rate in the prophylactic trial (92 versus 90%, RR 1.03, 95% CI 0.94-1.12) or in the therapeutic trial (81 versus 85%, RR 0.96, 95% CI 0.79-1.16). No effect of cervical occlusion was found in terms of gestational age at delivery and number of days the neonate spent in the NICU. Cervical occlusion was associated with no harm. CONCLUSIONS: Cervical occlusion with cerclage had no significant additional effect.