Objectives: Passive fit between restorations and implants is considered a significant factor in preventing mechanical and biologic failures in implant dentistry. Since impression taking and master cast fabrication have been shown to cause substantial errors, it was the aim of this study to evaluate the effect of intraoral scanning on the passivity of fit of three-unit cement-retained implant-supported fixed dental prostheses (FDPs). Method and Materials: Conventional and optical impressions (n = 10) of an in vitro situation with two implants in the mandibular left quadrant were made. Based on the scans, 10 FDP frameworks were CAD/CAM fabricated from zirconia ceramic; stereolithographic casts were also made. On both the stereolithographic casts and the casts obtained from the conventional impressions, 10 frameworks were waxed and cast in dental training alloy. Subsequently, all samples were cemented on the in vitro model, while the strains emerging were recorded for statistical analysis (MANOVA with Pillai trace; ß = .05). Results: The restorations fabricated on the stereolithographic casts showed significantly lower levels of strain than the conventionally fabricated ones (P = .013). No significant difference could be found between conventionally and CAD/CAM-fabricated frameworks (P = .080). Conclusion: Intraoral digitization of dental implants appears to be at least as precise as conventional impression taking and master cast fabrication using prefabricated transfer components and laboratory analogs. Schlagwörter: CAD/CAM, impression, passive fit, strain development