The aim of this study was to evaluate the vertical marginal gap of sintered gold copings and metal-ceramic crowns with different finish line preparations: a beveled round shoulder (BRS) and a beveled long chamfer (BLC), testing the null hypotheses that there are no differences in marginal gap regardless of finish line and phase of restoration (coping or crown). Stainless steel master models were fabricated to simulate tooth preparation for metal-ceramic crowns with different finish lines (BRS and BLC). Ten dies were obtained from each model. Preparations were coated with 2 layers of spacer to 1 mm from the margin. Sintercast gold copings were prepared, sintered and adjusted to the dies. The copings (n=10) were placed onto the master model and the marginal gap was measured in 24 equidistant points using optical microscopy (X230). An opaque and two body ceramic layers were subsequently applied to the copings and the same measuring procedure was performed for the crowns. The data were analyzed statistically using paired and unpaired Student's t-test (α=0.05). Mean marginal gap values (μm) for the copings and crowns were, respectively: 113.6 and 117.1 for the BRS; and 58.2 and 74.3 for the BLC preparation. Significantly greater marginal gaps (p=0.0307) were found for restorations with BRS than with BLC finish line, which also showed statistically significant differences in the gap size between coping and crown (p=0.001). In conclusion, marginal gap is influenced by ceramic application on copings fabricated on BLC preparation, and greater marginal gaps were found for restorations with BRS finish line, rejecting the experimental null hypotheses.