Objective: The objective of the present study was to compare and evaluate the accuracy of three-dimensional (3D) image data acquired from cone beam computed tomography (CBCT) and a dental scanner using 3D software. Materials and methods: After selecting the full-arch forms of the maxilla and mandible as the master cast, the master cast was scanned via a high-precision optical scanner for use as master model data. The model was scanned 12 times each using CBCT and a dental scanner. Scanned data were superimposed onto the master cast data for evaluation of accuracy and repeatability. Results: Although significant differences in both accuracy and repeatability were seen between CBCT and dental scanner (P 0.05), repeatability of the maxillary arch showed little difference, with CBCT and scanner having values of 17 ± 2 µm and 22 ± 5 µm, respectively. Meanwhile, repeatability of the mandibular arch with CBCT and scanner was 15 ± 0 µm and 19 ± 3 µm, respectively. Since good repeatability was shown, this demonstrated that data can be stably acquired. Conclusions: The present study demonstrated the feasibility of using a dental scanner to create a digital model as a substitute for a plaster model for use in orthodontic diagnosis and device fabrication. Palabras clave: cone beam computed tomography, dental scanner, accuracy, repeatability, digital model, 3D software