Background: This in-vivo study aimed to evaluate the marginal fit of three-unit monolithic zirconia fixed partial dentures (FPDs) using conventional and scannable polyvinyl siloxane impression materials because of the great importance of marginal accuracy. Methods: Ten patients were selected to replace the lateral teeth with a three-unit monolithic zirconia bridge. In the identical session, for each patient, as in the first group, an impression was made with a two-step putty-wash technique using scannable polyvinyl siloxane material (BONASCAN; DMP, Greece), and for the second group, an impression of conventional putty-wash polyvinyl siloxane was taken (BONASIL A⁺ Putty; DMP, Greece). The marginal discrepancy was measured once the replicas were cut perpendicularly within the buccolingual and mesiodistal directions. An Independent t-test was employed for data analyses (P < 0.05).Results: The marginal discrepancy in a conventional method for central abutment in four regions and canine abutment in mid-buccal, mid-lingual, and mid-distal was higher than in the scannable method but marginal discrepancies were not significant (P>0.05). However, the marginal discrepancy for mid-mesial in the scannable method was higher than in the conventional method, but it was not significant (P>0.05). Conclusions: FPDs fabricated from both scannable impression materials and conventional impression materials were not superior to each other in marginal fit by evaluation using the replica technique.