Background Coarctation of the aorta (CoA) is the most common undiagnosed congenital heart defect during prenatal screening. High false positive and false negative rates seriously affect prenatal consultation and postnatal management. The objective of the study was to assess the utility of various measurements to predict prenatal CoA and to derive a diagnostic algorithm. Methods One hundred and fifty-four fetuses with suspected CoA who presented at Fuwai Hospital between December 2017 and August 2021 were enrolled and divided into confirmed CoA cases (47) and false-positive cases (107), according to their postnatal outcomes. Fetal echocardiographic measurements included the cardiac chambers sizes, valve annular diameters, and ascending aorta and pulmonary artery dimensions. The corresponding gestational age-related Z-scores were collected. The diameter of the transverse aortic arch and the aortic isthmus were measured. The TAO-DAO angle between the transverse aortic arch and isthmus was also measured. The receiver operating characteristic curve determined the predictive capability of each diagnostic parameter, and the kappa test determined the diagnostic accuracy of the proposed model.Results Confirmed CoA cases had thinner transverse arches (1.92 ± 0.32 mm vs. 3.06 ± 0.67 mm, P = 0.0001), lower Z-scores of the isthmus (-8.97 ± 1.45 vs. -5.65 ± 1.60, P = 0.0001), smaller TAO-DAO angles (105.54 ± 11.51° vs. 125.29 ± 8.97°, P = 0.0001) and larger distance between the left subclavian artery and left common carotid artery (4.45 ± 1.75 mm vs. 2.74 ± 1.07 mm, P = 0.0001) than the false-positive cases. The area under the curve was 0.947 (95% CI 0.91–0.98) for the TAO-DAO angle, 0.942 (95% CI 0.91–0.98) for the transverse arch diameter, and 0.937 (95% CI 0.90–0.98) for the Z-score of the isthmus. Three important echocardiographic indices were combined to derive a three-step echocardiographic diagnostic model. The kappa test showed that the model’s diagnostic accuracy was consistent with postnatal outcomes (kappa value 0.936, P = 0.0001).Conclusions The three step approach included the three most useful measurements and the five additional indices with appropriate cut-off values. The algorithm optimizes the diagnosis of fetal CoA and is clinically feasible.Trial registration: retrospectively registered.

No abstract