OBJECTIVE To calculate the precision of the implant analog position in digital models generated from different computer-assisted design and computer-assisted manufacturing (CAD-CAM) systems compared to gypsum models acquired from conventional implant impressions. MATERIALS AND METHODS In five patients in need of a single implant crown, a within-subject comparison was performed applying four different manufacturing processes for the implant model. Each implant was scanned with three different intraoral scanners: iTero Cadent (ITE), Lava True Definition (LTD), and Trios 3Shape (TRI). All digital implant models were fabricated using the corresponding certified CAD-CAM workflow. In addition, a conventional impression was taken (CON) and a gypsum model fabricated. Three consecutive impressions were acquired with each impression system. Following fabrication, all implant models were scanned. The datasets were aligned by a repeated best-fit algorithm and the precision for the implant analog and the adjacent teeth was measured. The precision served as a measure for reproducibility. RESULTS Mean precision values of the implant analog in the digital models were 57.2 ± 32.6 µm (ITE), 88.6 ± 46.0 µm (TRI), and 176.7 ± 120.4 µm (LTD). Group CON (32.7 ± 11.6 µm) demonstrated a statistically significantly lower mean precision value for the implant position in the implant model as compared to all other groups representing a high reproducibility. The mean precision values for the reference ranged between 31.4 ± 3.5 µm (TRI) and 39.5 ± 16.5 µm (ITE). No statistical significant difference was calculated between the four treatment groups. CONCLUSIONS The conventional implant model represented the greatest reproducibility of the implant position. Digital implant models demonstrated less precision compared to the conventional workflow.