Purpose: The aim of this study was to compare the marginal adaptation of metal-ceramic crowns made with 3 different metal substructures. Materials and Methods: Seventy-five gypsum dies were fabricated from a master steel die. They were randomly divided into 3 groups of 25 dies each. A metal coping was prepared on each die using a different technique for each group: composite alloy, electroforming, and cast high-noble alloy. Ceramic was baked on all of the samples following the same protocol and simulating the shape and size of a maxillary central incisor. All specimens were evaluated under a stereomicroscope at 200! magnification. The marginal gap was measured at 8 different points on each specimen before and after ceramic application and also on the master steel die. A total of 3,600 measurements was evaluated. The results were statistically elaborated by means of a t test and one-way analysis of variance (ANOVA). Results: The ANOVA revealed a significant difference in mean marginal gap before and after ceramic application within each group. Electroforming (mean 32 µm) and high-noble alloy (mean 31 µm) showed better marginal fit after ceramic application and on the master steel die compared to composite alloy (mean 68 µm; P 0.0001). Conclusion: Marginal gaps are influenced by the application of ceramic and by the alloy used for the substructure, and the values obtained were within the limits of clinical acceptability (70 µm).