Objective: To examine the value of the cervical length (CL) measurement at 24-30 gestational weeks in the prediction of spontaneous preterm delivery (SPD) between 30 and 34 weeks (SPD34) and between 34 and 37 weeks (SPD37). Methods: We performed a prospective cross-sectional study. CL was measured once by transvaginal ultrasound examination between 24 and 30 weeks. Results: The study sample consisted of 1,180 low-risk singleton pregnancies. 10 women (0.85%) had a SPD34 and 60 (5.08%) had a SPD37. CL was shorter (p < 0.001) in the women who had a SPD34 (median 11 mm) compared to the women who delivered after 34 weeks (median 31 mm). CL was shorter (p < 0.001) in the women who had a SPD37 (median 22 mm) compared to the women who delivered after 37 weeks (median 31 mm). CL predicted SPD34 (OR = 0.837, R2 = 0.2768, AUC = 0.9406, p < 0.001) and SPD37 (OR = 0.907, R2 = 0.1085, AUC = 0.7584, p < 0.001). The model achieved a sensitivity of 70.0 and 38.3% for 10% false-positive rate for SPD34 and SPD37, respectively. Conclusions: CL after 24 weeks is significantly shorter in women destined to have a SPD. In low-risk singleton pregnancies CL performs very well in predicting SPD34 and adequately in predicting SPD37.