Occlusal collisions of articulated intraoral digital scans can be corrected by intraoral scanners (IOSs) or dental design software programs. However, the influence of these corrections on the accuracy of maxillomandibular relationship is unclear. The purpose of this clinical investigation was to measure the effect of occlusal collision corrections completed by the IOSs or dental design software programs on the trueness and precision of maxillomandibular relationship. Casts of a participant mounted on an articulator were digitized (T710). The experimental scans were obtained by using 2 IOSs: TRIOS4 and i700. The intraoral digital scans of the maxillary and mandibular arches were obtained and duplicated 15 times. For each duplicated pair of scans, a bilateral virtual occlusal record was acquired. Articulated specimens were duplicated and assigned into 2 groups: IOS-not corrected and IOS corrected (n=15). In the IOS-not corrected groups, the IOS software program postprocessed the scans maintaining the occlusal collisions, while in the IOS-corrected groups, the IOS software program eliminated the occlusal collisions. All articulated specimens were imported into a computer-aided design (CAD) program (DentalCAD). Three subgroups were developed based on the CAD correction: CAD-no change, trimming, or opening the vertical dimension. Thirty-six interlandmark distances were measured on the reference and each experimental scan to compute discrepancies by using a software program (Geomagic Wrap). Root mean square (RMS) was selected to compute the cast modifications performed in the trimming subgroups. Trueness was examined using 2-way ANOVA and pairwise comparison Tukey tests (α=.05). Precision was evaluated with the Levene test (α=.05). The IOS (P<.001), the program (P<.001), and their interaction (P<.001) impacted the trueness of the maxillomandibular relationship. The i700 obtained higher trueness than the TRIOS4 (P<.001). The IOS-not corrected-CAD-no-changes and IOS-not-corrected-trimming subgroups obtained the lowest trueness (P<.001), while the IOS-corrected-CAD-no-changes, IOS-corrected-trimming, and IOS-corrected-opening subgroups showed the highest trueness (P<.001). No significant differences in precision were found (P<.001). Furthermore, significant RMS differences were found (P<.001), with a significant interaction between Group×Subgroup (P<.001). The IOS-not corrected-trimmed subgroups obtained a significantly higher RMS error discrepancy than IOS-corrected-trimmed subgroups (P<.001). The Levene test showed a significant discrepancy in the RMS precision among IOSs across subgroups (P<.001). The trueness of the maxillomandibular relationship was influenced by the scanner and program used to correct occlusal collisions. Better trueness was obtained when the occlusal collisions were adjusted by the IOS program compared with the CAD program. Precision was not significantly influenced by the occlusal collision correction method. CAD corrections did not improve the results of the IOS software. Additionally, the trimming option caused volumetric changes on the occlusal surfaces of intraoral scans.