Model analysis remains an important part of orthodontic diagnosis. Several techniques used for measuring dental models. Coordinate Measuring Machine (CMM) have been used recently. Objectives: The purpose of this study was to evaluate the reliability of CMM in measuring dental models. Method: Forty sets of normal occlusion study casts comprised the sample. The arch dimension values (canine width, molar width and arch depth) which were measured using different methods (manual, digitization, and CMM) were compared. The corresponding X, Y coordinates of dental landmarks between measurements obtained from digitization and CMM were also compared. Results: Both canine and molar widths were significantly different between measurements from digitization and manual method and between measurements from CMM and manual method at p<0.001. However, the mean differences were very small (range of mean differences of canine and molar widths was 0.08-0.27 and 0.18-0.22mm, respectively). In the comparison of arch depth, there was statistically significant difference of maxillary arch depth between the measurements from the digitization and manual methods at p<0.001. The mean difference was 0.62mm which was clinically significant. However, no significant difference of arch depth was found between the CMM and manual measurements. There was no statistically significant difference for the X-coordinate of each landmark both on the maxillary and mandibular dental casts between the measurements from digitization and CMM. Statistical significant differences of Y-coordinates were found in twelve and ten of the eighteen maxillary and mandibular dental landmarks, respectively (p<0.001). Conclusions: Vernier gauge measurements on dental casts produce the most accurate, reliable, and reproducible results. While dental casts cannot be accurately reproduced on photocopies due to the curve of Spee, CMM can apply for measuring arch dimensions with highly accurate, reliable, and reproducible results.