Statement of problemDigital light processing (DLP), continuous liquid interface printing (CLIP), and stereolithography (SLA) technologies enable 3-dimensional (3D) printing of surgical guides. However, how their accuracy compares and how accuracy may affect subsequent steps in guided surgery is unclear. PurposeThe purpose of this in vitro study was to investigate the fabrication and seating accuracy of surgical guides printed by using DLP, SLA, and CLIP technologies and evaluate the positional deviation of the osteotomy site and placed implant compared with the digital implant plan. Material and methodsTwenty-one polyurethane models were divided into 3 groups and used to plan implants and design surgical guides. The guides were fabricated by using DLP, SLA, or CLIP 3D printers (n=7) and scanned, and the scan file was compared with the digital design file to analyze the fabrication accuracy at the intaglio and overall external surfaces using root mean square (RMS) values. The triple scan protocol was used to evaluate the seating accuracy of the guides on their respective models. Osteotomies were prepared on models by using the guides followed by a microcomputed tomography image of each osteotomy. The implants were placed through the guides, the scan bodies were tightened to implants, and the models were scanned to obtain the images of placed implant position. Osteotomy and placed implant images were used to calculate the entry point, apex, and long axis deviations from the planned implant position with a software program. A 2-way repeated-measures ANOVA of the RMS data was used to analyze printing and seating trueness, and homogeneity of variance analyses were used at each surface for precision. A 3-way repeated-measures ANOVA was used to analyze distance deviations over the stages (osteotomy and final implant) and locations studied, and a 2-way repeated-measures ANOVA was used for angular deviations. Homogeneity of variance analyses were performed for precision (α=.05). ResultsThe 3D printer type significantly affected the trueness of the guide at the intaglio surface (P<.001). SLA guides had the lowest mean RMS (59.04 μm) for intaglio surface, while CLIP had the highest mean RMS (117.14 μm). Guides from all 3D printers had low variability among measured deviations and therefore were similarly precise. The seating accuracy of SLA and DLP guides was not significantly different, but both had lower mean RMS values than CLIP (P=.003 for SLA, P=.014 for DLP). There were no significant interactions between the stage of surgery, the printer type, or the location of implant deviation (P=.734). Only the location of deviation (cervical versus apical) had a significant effect on distance deviations (P<.001). The printer type, stage of surgery, and their interaction did not significantly affect angular deviations (P=.41). ConclusionsThe 3D printing technology affected printing trueness. The intaglio surface trueness was higher with SLA and overall trueness was higher with the CLIP printer. The precision of all guides was similarly high. Guides from SLA and DLP printers had more accurate seating than those from CLIP. Higher deviations were observed at the apex; however, osteotomy and final implant position did not significantly differ from the digitally planned position.