The aim of the study was to determine potential deviation in measurements of commonly used orthodontic variables, including intermolar width (IMW), intercanine width (ICW), and arch perimeter (AP), when measured on three-dimensional images compared to measurements performed on plaster models obtained from alginate impressions using a hand-held digital vernier caliper. A total of 120 sets of models were divided into four groups: plaster models made from an alginate impression, digitized plaster models made from an alginate and silicone impression, and digital models obtained via intraoral scan. The repeated measurements analysis of variance (ANOVA) with Sidak post-hoc test was applied to test differences in measurements between the four methods. The results show better agreement for intermolar and intercanine width (r = 0.468–0.967) and smaller measurement error (ME = 0.53–1.07) compared to arch length (r = 0.422–0.477; ME = 2.04–2.89). Better agreement levels were more frequently observed between plaster models with intraoral scans than between digitized plaster models. The deviations in intermolar and intercanine width measurements between the hand-held digital vernier caliper and digital analyses were found to be clinically insignificant. The observed difference in arch perimeter suggests that curved measurements may not be clinically acceptable.