The purpose of the present study was to examine the impact of scanning strategy on trueness and precision of the impression acquired from an intraoral scanner. Fifteen complete-arch, mandibular, post-orthodontic treatment casts were scanned with a laboratory scanner (Identica SE 3D, Medit) as the gold standard, and with an intraoral scanner (i500 Medit) following three different paths of the scanning head over the arch (scanning strategies A, B, and C). The hand scans were performed twice by one examiner and repeated by a second examiner, resulting in 180 triangular mesh surfaces (digital casts). The meshes were superimposed on the gold standards using the Viewbox 4 software. The closest distances between the meshes were computed and trueness and precision were evaluated using a General Linear Model. An interaction was found among the examiner and strategy. The accuracy of complete-arch impressions was affected by the scanning strategy; the manufacturer’s recommended strategy (A) was statistically significantly better (p < 0.05) than B and C, which were similar. An average accuracy below 50 μm, which is clinically acceptable in most orthodontic procedures, was achieved with all the examined scanning strategies.