This study aimed to evaluate the accuracy of three-dimensional (3D) cephalometric analysis in virtual dentoskeletal models by comparing it with cone-beam computed tomography (CBCT) 3D models and identifying any significant differences. The virtual dentoskeletal models are created by integrating CBCT with digital dental models. The dental casts are digitized using the structure-from-motion (SfM) photogrammetry method to create digital 3D models. The research included seven patients who underwent orthognathic surgery. The 3D cephalometric analysis was calculated using 27 cephalometric landmarks and 18 measurements (14 angles and four linear). Statistical analyses included paired sample t-tests and Bland-Altman plots. Statistical analysis showed that the differences of the linear and angular measurements are statistically significant for differences like U1/NA, U1-NA, U1/SN, L1-NB, U1/L1, overjet, and overbite, with the p-value < 0.05. The mean differences ranged from -1.26° to 1.66° for angular measurements and 0.057mm to 0.329mm for linear measurements. Notably, the agreement interval shows a substantial difference for angular measurements (-4.38 to 4.38) in the Bland-Altman plot, while a minor difference was noted for linear measurements (-1.34 to 0.91). The differences were evaluated clinically by comparing them to an acceptable clinical boundary of 0.5 mm. Integrating two models to create a virtual dentoskeletal model is a robust technique that enhances the precision of the dental region in the resulting 3D model. In addition to improving the accuracy of 3D cephalometric analysis for orthognathic surgery planning and orthodontic treatment, this innovation holds promise as a valuable tool for dentists and orthodontists. This technology has the potential to enhance patient care in dentistry.