Purpose This study aimed to evaluate the trueness and precision of complete-arch models printed with three-dimensional printers via three different printing technologies.

Methods An arch-shaped master model was designed using software (RapidForm XOR2, 3D Systems Inc., USA), and the digital master model was printed 10 times with three-dimensional printers using stereolithography (SLA), direct light processing (DLP), and Polyjet technology (n = 30). The printed models were then scanned with an industrial scanner to create the respective digital models. All digital models were compared with the master model, and an evaluation of the trueness was performed by model superimposition with Geomagic Control software (3D Systems, Rock Hill, SC, USA). Precision was determined for each case by superimposing some combination of the 10 datasets in each group.

Results The trueness of the printed models was 46.2 µm for the DLP printer, 51.6 µm for the SLA printer, and 58.6 µm for the Polyjet printer. The DLP models were significantly better than the Polyjet models (p = .005). However, the Polyjet models (30.4 µm) were more precise than the SLA (37.6 µm) and DLP (43.6 µm) models (p < .001, p = .016). Furthermore, the SLA (11.8 µm) was the most accurate printer in the Z-direction (p = .016, p = .002).

Conclusions The 3D printing technologies showed significant differences in the precision and trueness of complete-arch measurements. Although DLP was more accurate other tested 3D printers, the accuracy of all 3D printed models was within clinical tolerance, and they were clinically acceptable and could be used for the production of fixed restorations.