ObjectiveThis in vitro study aimed to determine whether a newly designed arcuate scan body can improve intraoral scanning accuracy for implant rehabilitation of edentulous jaws. Material and methodsA master model containing 4 implant abutment replicas was fabricated and digitized with different scan bodies using an intraoral scanner. Four types of scan bodies were evaluated: original scan bodies (group OS), computer-aided design and computer-aided manufacturing (CAD/CAM) scan bodies without extension (group CS), CAD/CAM scan bodies with straight extension (group CSS), and CAD/CAM scan bodies with arcuate extension (group CSA). Conventional splinted open-tray impressions (group CI) were used as controls. The master model and the poured casts were digitized using a laboratory scanner. Impressions were repeated 10 times each in 5 groups. Scans in standard tessellation language format were exported to reverse engineering software and root mean square (RMS) values were used for trueness and precision assessments. In each group, 45 RMS values were acquired for precision evaluation and 10 RMS values were obtained for trueness assessment. Statistical evaluation was performed with the Kruskal–Wallis test and Dunn–Bonferroni test (α = 0.05). ResultsThe median trueness values were 41.40, 55.95, 39.80, 39.75, and 22.30 μm for group OS, CS, CSS, CSA, and CI, respectively. CI showed better trueness than OS (P = .020), CS (P < .001), and CSS (P = .035). The median precisions for group OS, CS, CSS, CSA, and CI were 47.40, 51.50, 43.90, 25.20, and 24.60 μm. respectively. The precision of CSA and CI were higher than OS (P < .001), CS (P < .001), and CSS (P < .001). Between CI and CSA, there was no significant difference (P = 1.000). ConclusionsFor full-arch implant rehabilitation, the scan body with arcuate extension could improve the intraoral scanning precision and showed similar 3-dimensional discrepancy compared to conventional splinted open-tray impressions.