Background This study aimed to compare the precision and trueness of digital impressions of the edentulous arch made with different scanners to conventional physical impressions. Methods A total of 40 impressions of a completely edentulous maxillary arch model (n = 10) were made using different digital impressions with an extraoral scanner, E3 3Shape desktop scanner, as the reference scan, intraoral scanner (TRIOS IOS, and Medit IOS) and Vinyl Polysiloxane impressions (VPS) impression using a Computer-Aided Design and Computer-Aided Manufacturing (CAD-CAM) custom tray. The VPS impression was scanned with the desktop scanner to produce standard tessellation language (STL) files for comparison with the digital impressions made by the Desktop and intraoral scanners. The STL files were super-imposed to a desktop scan and to each other with the same group using Geomagic Control X Software to assess the trueness and precision, respectively. A t-test was conducted for statistical analysis with a significance level of 0.05. Results The overall trueness, Medit had the highest deviation compared to the VPS and TRIOS groups with a P value of 0.0013 and <0.0001, respectively. In terms of overall precision, TRIOS had a lower deviation than the VPS group, with a P value of 0.0002. The TRIOS and Medit groups had statistically comparable results. The desktop scanner showed the highest precision in digitizing completely edentulous cases, followed by the TRIOS scanner. The Medit scanner’s trueness had the highest deviation compared to the VPS and TRIOS groups.