PURPOSETo evaluate the fit of a crown produced based on a 3D printed model and to investigate its clinical applicability.MATERIALS AND METHODSA master die was fabricated with epoxy. Stone dies were fabricated from conventional impressions (Conventional stone die group: CS, n=10). Digital virtual dies were fabricated by making digital impressions (Digital Virtual die group: VD, n=10). 3D data obtained from the digital impression was used to fabricate 3D printed models (DLP die group: DD, n=10, PolyJet die group: PD, n=10). A total of 40 crowns were fabricated with a milling machine, based on CS, VD, DD and PD. The inner surface of all crowns was superimposed with the master die files by the “Best-fit alignment” method using the analysis software. One-way and 2-way ANOVA were performed to identify significant differences among the groups and areas and their interactive effects (α=.05). Tukey's HSD was used for post-hoc analysis.RESULTSOne-way ANOVA results revealed a significantly higher RMS value in the 3D printed models (DD and PD) than in the CS and DV. The RMS values of PD were the largest among the four groups. Statistically significant differences among groups (P<.001) and between areas (P<.001) were further revealed by 2-way ANOVA.CONCLUSIONAlthough the fit of crowns fabricated based on the 3D printed models (DD and PD) was inferior to that of crowns prepared with CS and DV, the values of all four groups were within the clinically acceptable range (<120 µm).