Objectives To collect normative data and test the feasibility and reproducibility of measurement of the maxilla-nasion-mandible (MNM) angle between 16 and 36 weeks' gestation and its diagnostic ability in a group of pathological cases. Methods The MNM angle is defined as the angle between the intersection of the maxilla-nasion and mandible-nasion lines in the exact mid-sagittal plane. After assessing reproducibility, the MNM angle was measured in 3D volumes in 241 fetuses cross-sectionally and in 11 fetuses longitudinally. The MNM angle was then tested in 18 pathological cases with facial malformations or syndromes with specific facial features. Results The MNM angle could be measured in 92.3% of normal fetuses. Intra- and interobserver intraclass correlation coefficient (ICC) variability was 0.92 and 0.81, respectively. The difference between paired measurements performed by one or two observers was less than 2.5° and 3.6°, respectively in 95% of the cases. The mean MNM angle was 13.5° and did not change significantly during pregnancy (r = - 0.08, P = 0.25). The MNM angle was above the 95th centile in all cases of retrognathia and maxillary alveolar ridge interruption. The MNM angle was below the 5th centile in Apert syndrome, thanatophoric dysplasia and in two of the three Down syndrome cases. Conclusions The feasibility and reproducibility of measurement of the MNM angle is good. The MNM angle can be used to evaluate the convexity of the fetal profile by enabling an objective assessment of the anteroposterior relationship of the jaws and it may therefore be of help in the diagnosis of retrognathia, maxillary alveolar ridge interruption and flat profile. Copyright © 2011 ISUOG. Published by John Wiley & Sons, Ltd.