Women with multiple gestations often deliver at less than 37 weeks, increasing the risks of perinatal morbidity and mortality. Treatment with a pessary might prevent preterm birth by changing the inclination of the cervical canal and preventing premature dilatation of the cervix, rupture of the membranes, and deterioration or loss of the cervical mucous plug. This multicenter, open-label, randomized controlled trial at 40 hospitals was performed to determine whether a cervical pessary could prevent poor perinatal outcomes in parturients with a multiple pregnancy.At 12 to 20 weeks' gestation, patients were assigned to the pessary or control group; cervical length was measured at 16 to 22 weeks. Women in the study group received an Arabin pessary at 16 to 20 weeks, which was removed in week 36 or in case of preterm premature rupture of the membranes, active vaginal bleeding, other signs of preterm labor, or patient discomfort. Obstetric care was otherwise similar in the 2 groups. The primary outcome was a composite of poor perinatal outcomes, including stillbirth, preventricular leukomalacia, severe respiratory distress syndrome, bronchopulmonary dysplasia, intraventricular hemorrhage, necrotizing enterocolitis, sepsis, and neonatal death within 6 weeks after the anticipated term date. Secondary outcomes were time to delivery, preterm birth at less than 32 and less than 37 weeks, days in the neonatal intensive care unit, days of maternal admission for preterm labor, and maternal morbidity.A total of 401 of 403 women in the pessary group and 407 of 410 in the control group completed the study. Five women in the pessary group had a surgical cerclage; 1 patient died, and the others delivered at 21.6 to 36.7 weeks, with 3 having poor perinatal outcomes. No patients in the control group had a cerclage. Vaginal discharge occurred in 104 women (26%) in the pessary group and in none of the controls. The pessary was removed from 57 women (14%) at less than 28 weeks and from 22 women (5%) at 28 to 32 weeks; 7 and 13, respectively, delivered within 48 hours of removal. At 32 to 36 weeks, the pessary was removed from 107 women; 70 delivered within 48 hours. The most common reasons for pessary removal in these women were preterm premature rupture of the membranes, vaginal bleeding, contractions, and induction of labor. The composite poor perinatal outcome occurred in 53 women (13%) and 55 women (14%) in the pessary and control groups, respectively (relative risk (RR), 0.98; 95% confidence interval, 0.69-1.39). Ten stillbirths (2%) occurred in each group. The other conditions within the composite outcome did not differ between the 2 groups. In the pessary and control groups, 16 and 18 infants, respectively, died before discharge. The groups were similar in median gestational age at delivery; frequencies of delivery at less than 28, less than 32, and less than 37 weeks; and frequency and length of neonatal intensive care unit admission. In the women with a cervical length of less than 25th percentile (<38 mm), the pessary significantly reduced the frequency of poor perinatal outcomes and very preterm delivery.A cervical pessary does not necessarily prevent poor perinatal outcomes or preterm birth in all women with a multiple gestation. Although the pessary had a positive effect in women with a twin pregnancy and a short cervix, these results should be confirmed in additional prospective studies. The safety and low cost of the pessary should be considered when counseling a parturient with a multiple pregnancy and short cervix.