Introduction: To compare the effect of a policy of screening for spontaneous preterm delivery (SPD) by transvaginal cervical length (CL) measurement versus a no screening policy in the prevention of severe prematurity. Methods: Retrospective study on low risk singleton pregnancies examined at 20-24 weeks. Two cohorts one with SPD screening and the other without screening were matched using propensity analysis to create the study groups. Women with short CL were treated with vaginal progesterone and/or cervical cerclage/pessary. The outcomes examined were SPD<32 weeks (SPD 32) and SPD between 20 and 32 weeks (SPD 20-32). Results: Screening for SPD was associated with a significant reduction in the rate of SPD at less than 32 weeks (0.3% vs 0.8%, p=0.001 in the screened and no screened pregnancies respectively) and in the rate of SPD 20-32 (0.3% vs 0.9%, p=0.005 in the screened and no screened pregnancies respectively). After adjusting for maternal age, parity, body mass index, smoking and mode of conception, the screening group had significantly lower hazard for SPD 20-32 (HR=0.36, 95% CI: 0.18-0.75, p=0.006) and SPD32 (HR=0.39, 95% CI: 0.19-0.82, p=0.013). Conclusion: Screening for SPD by transvaginal CL measurement in mid pregnancy may reduce the incidence of severe prematurity in low risk singleton pregnancies.