Objective Cervical elastography has been used in pregnant women to diagnose preterm births. However, there is a variability in the measured elasticity parameters and imaging mode used. We evaluated the precision of cervical elastography in identifying preterm births. Methods Extensive and methodical searches were made in the databases such as Scopus, Embase, Cochrane Library, PubMed Central, Medline, ScienceDirect, and Google Scholar from the inception until November 2022, for studies that report diagnostic accuracy of cervical elastography for preterm deliveries in antenatal women. Results The pooled sensitivity and specificity value of cervical elastography for preterm deliveries were 82% (95%CI: 73%–89%) and 77% (95%CI: 64%–86%), respectively with area under curve (AUC) of 0.87 (95%CI: 0.72–0.95). The diagnostic odds ratio (DOR) was 15 (95%CI: 8–28), positive likelihood ratio (LRP) was 3.5 (95%CI: 2.3–5.5) and negative likelihood ratio LRN was 0.23 (0.16–0.34). Pooled sensitivity and specificity of shear wave elastography was 88% and 71%, respectively. Pooled sensitivity and specificity of strain elastography was 80% and 79%, respectively. Heterogeneity was significant, as indicated by chi-square test and an I2 statistic of over 75. Conclusions Cervical elastography can be used for predicting preterm deliveries with moderate to high level of accuracy.