To investigate the effects of scan body (SB) angulation and geometric attachments on the deviations of complete-arch digital implant impressions obtained using intraoral scanners (IOSs), in relation to the gold-standard desktop scanner. Two intraoral scanners (iTero and Omnicam), two scan body angulations (0 degrees and 30 degrees), and geometric attachments (GAs) for the SBs were investigated. SBs were attached to an edentulous maxillary cast with the following implant analogs: parallel 0-degree at positions #13, #23, and #26, and 30-degree posteriorly tilted at position #16. The cast was digitized using a reference desktop scanner, followed by ten consecutive digital scans using each IOS (five scans with GAs and five without GAs, n = 20). Meshes obtained from the IOS scans were superimposed on the master reference mesh. Linear distance and 3D surface deviations from the reference mesh were calculated. A three-way multivariate analysis of variance (MANOVA) was employed to assess the effects of angulation, IOS type, and GAs on the combined dependent variables (α = 0.05). No significant three-way interaction was observed between IOS type, SB angulation, and GAs for combined deviations from the reference scan (p = 0.56). Simple main effect analysis revealed that iTero exhibited significantly lower mean 3D surface and linear deviations than Omnicam (p<0.05). Additionally, the use of GAs significantly reduced deviations (p<0.001), with mean deviation reductions for both scanners ranging from 26-33 micrometers (μm) for 3D deviations and 15-21μm for linear distance deviations. No differences were found between angled and parallel SBs regarding 3D surface and linear distance deviations (p≥0.05). iTero demonstrated significantly smaller deviations, and the use of GAs led to significantly reduced distance and 3D surface deviations. SB angulations did not impact scan deviations. This article is protected by copyright. All rights reserved.