Objectives: Little information is available on the impact of different scan strategies on the accuracy of full-arch scans with intraoral scanners. The aim of this in-vitro study was to investigate the trueness and precision of full-arch maxillary digital impressions comparing three scan strategies. Method and Materials: Three scan strategies (A, B, and C) were applied each five times on one single model (A, first buccal surfaces, return from occlusal-palatal; B, first occlusal-palatal, return buccal; C, S-type one-way). The TRIOS Pod scanner (3shape, Copenhagen, Denmark) with a color detector was used for these digital impressions. A cast of a maxillary dentate jaw was fabricated and scanned with an industrial reference scanner. This full-arch data record was digitally superimposed with the test scans (trueness) and within-group comparison was performed for each group (precision). The values within the 90/10 percentiles from the digital superimposition were used for calculation and group comparisons with nonparametric tests (ANOVA, post-hoc Bonferroni). Results: The trueness (mean ± standard deviation) was 17.9 ± 16.4 μm for scan strategy A, 17.1 ± 13.7 μm for B, and 26.8 ± 14.7 μm for C without statistically significant difference. The precision was lowest for scan strategy A (35.0 ± 51.1 μm) and significantly different to B (7.9 ± 5.6 μm) and C (8.5 ± 6.3 μm). Conclusions: Scan strategy B may be recommended as it provides the highest trueness and precision in full-arch scans and therefore minimizes inaccuracies in the final reconstruction. Keywords: accuracy, digital impression, intraoral scan, precision, strategy, trueness