Background: To test the hypothesis that transparent matrices result in more continuous margins of bulk-fill composite (BFC) restorations than metal matrices. Methods: Forty standardized MOD cavities in human molars with cervical margins in enamel and dentin were created and randomly assigned to four restorative treatment protocols: conventional nanohybrid composite (NANO) restoration (Tetric EvoCeram, Ivoclar Vivadent, Schaan, Liechtenstein) with a metal matrix (NANO-METAL) versus transparent matrix (NANO-TRANS), and bulk-fill composite restoration (Tetric EvoCeram Bulk Fill, Ivoclar Vivadent, Schaan, Liechtenstein) with a metal matrix (BFC-METAL) versus transparent matrix (BFC-TRANS). After artificial aging (2500 thermal cycles), marginal quality was evaluated by scanning electron microscopy using the replica technique. Statistical analyses were performed using the Mann−Whitney U-test and Wilcoxon test. The level of significance was p < 0.05. Results: Metal matrices yielded significantly (p = 0.0011) more continuous margins (46.211%) than transparent matrices (27.073%). Differences in continuous margins between NANO (34.482%) and BFC (38.802%) were not significant (p = 0.56). Matrix type did not influence marginal gap formation in BFC (p = 0.27) but did in NANO restorations (p = 0.001). Conclusion: Metal matrices positively influence the marginal quality of class II composite restorations, especially in deep cavity areas. The bulk-fill composite seems to be less sensitive to the influence of factors such as light polymerization and matrix type.
Keywords: SEM; bulk-fill technique; centripetal technique; class II restoration; marginal gap formation; metal matrix; transparent matrix.