PURPOSE: The work presents outcomes of simulated maxillofacial skeleton osteotomies supported with cutting guides or one of two AR-based intraoperative navigation systems. MATERIAL AND METHODS: The series of osteotomies supported with a cutting guide, simple AR (sAR) or navigated AR (nAR) module were carried out on 15 (five per each method) skull models according to the virtual surgical plan (VSP). Each method was used to support 40 osteotomies involving the upper jaw and 40 involving the orbital region (16 osteotomies on each model). Postoperative computed tomography scans were fused with the VSP to analyse angular deviations from the planned cutting trajectory (°) and the deviations of labelled control points (mm). RESULTS: Guides provided the highest accuracy, with a mean osteotomy angular deviation of 3.73 ± 2.94° and a mean control point deviation of 1.30 ± 0.73 mm. Mean angular deviations for the sAR- and nAR-assisted osteotomies were 5.93 ± 5.12° and 6.75 ± 5.33°, and mean control point deviations amounted to 1.86 ± 0.88 mm and 1.97 ± 0.70 mm.