BackgroundIn this study we aimed at quantifying the possible errors which may occur when assessing specific reference planes and linear derivants on cephalometric radiographs traced manually and digitally. Furthermore, we have compared the precision of the tracings according to both the two- and three-dimensional (2D and 3D respectively) techniques and between clinicians.FindingsWe have obtained via cone beam computed tomography (CBCT) archive of the orthodontics department of the University of Milan 20 cone beam CT radiographs from which we have obtained 20 latero-lateral radiographs. Five independent clinicians referred to as A, B, C, D, E have been randomly selected to trace both radiographs maintaining the same working and lighting conditions to minimize the possibility of operator- and environment-dependent errors from occurring. The results have been statistically assessed by Student’s t test. The comparison of the data gathered from the tracings in 2D and 3D shows that certain measurements have statistically significant differences. Particularly, the difference in the measurements of the sagittal dimension of the mandible and the anterior and posterior nasal spines has resulted to be statistically significant. The results of the intra-operator comparison proved that the 3D technique is extremely precise.ConclusionOur study determines that the 3D technique allows to obtain more precise results and with several advantages when compared to the conventional technique such as a true representation of the anatomical structures, less risk of errors occurring due to clinician skills and absence of overlapping anatomical structures.