To verify predicted versus obtained surgical movements in 2-dimensional (2D) and 3-dimensional (3D) measurements and compare the equivalence between these methods. A retrospective observational study of bimaxillary orthognathic surgeries was performed. Postoperative cone-beam computed tomographic (CBCT) scans were superimposed on preoperative scans and a lateral cephalometric radiograph was generated from each CBCT scan. After identification of the sella, nasion, and upper central incisor tip landmarks on 2D and 3D images, actual and planned movements were compared by cephalometric measurements. One-sample t test was used to statistically evaluate results, with expected mean discrepancy values ranging from 0 to 2mm. Equivalence of 2D and 3D values was compared using paired t test. The final sample of 46 cases showed by 2D cephalometry that differences between actual and planned movements in the horizontal axis were statistically relevant for expected means of 0, 0.5, and 2mm without relevance for expected means of 1 and 1.5mm; vertical movements were statistically relevant for expected means of 0 and 0.5mm without relevance for expected means of 1, 1.5, and 2mm. For 3D cephalometry in the horizontal axis, there were statistically relevant differences for expected means of 0, 1.5, and 2mm without relevance for expected means of 0.5 and 1mm; vertical movements showed statistically relevant differences for expected means of 0, 0.5, 1.5 and 2mm without relevance for the expected mean of 1mm. Comparison of 2D and 3D values displayed statistical differences for the horizontal and vertical axes. Comparison of 2D and 3D surgical outcome assessments should be performed with caution because there seems to be a difference in acceptable levels of accuracy between these 2 methods of evaluation. Moreover, 3D accuracy studies should no longer rely on a 2-mm level of discrepancy but on a 1-mm level.