Résumé : Preterm delivery is strongly associated with neonatal mortality and morbidity. In some European countries, cervical examinations are used routinely during pregnancy to identify women at risk of preterm delivery. We sought to evaluate the efficacy and secondary effects of these routine cervical examinations. We did a randomised controlled trial in seven European countries, comparing two policies - namely, an attempt to do a cervical examination at every prenatal visit (2803 women) and avoidance of cervical examination if possible (2799). The median number of cervical examinations was 6 in the experimental group and 1 in the controls. There were 6.7% preterm (<37 weeks) deliveries in the experimental group and 6.4% in the control group risk ratio 1.05 [95% confidence interval 0.85-1.29]; non-significant). The low birthweight rate was 6.6% in the experimental group and 7.7% in the controls (non-significant). Premature rupture of membranes was not significantly more frequent in the experimental group (27.1% vs 26.5%). Our findings do not support the routine use of cervical examinations during pregnancy.