Objective: To investigate the cervical microleakage in Class II cavities restored with Sonicsys approx ceramic inserts and four resin-based materials. Method and Materials: Forty noncarious and crack-free mandibular third molars were used. These teeth were randomly assigned to four groups each containing 10 teeth. No control group was created. On each tooth, one mesial boxlike cavity was prepared using the active head Sonicsys approx no. 3. The cervical margin of each cavity was in enamel about 1 mm coronal to the cementoenamel junction. According to manufacturer's instructions, the prepared cavities were restored using a Sonicsys approx ceramic inserts no.3 and four resin-based materials as follows: group 1, Tetric Flow; group 2, Admira Flow; group 3, Nexus 2; group 4, X-Flow. After finishing and polishing, all specimens were stored in distilled water for 7 days at 37°C, thermocycled 1,000 cycles between 5° and 55°C, and stored for 24 hours in basic fuchsine 2%. All specimens were then embedded in clear acrylic resin and sectioned along a mesial-distal plane through the middle of the cervical margin. The cervical areas of the resulting sections were examined using an optical microscope to assess the dye penetration. The registered scores were analyzed using Kruskal-Wallis and Mann-Whitney U tests. Results: Microleakage was detected in each experimental group. Kruskal-Wallis test revealed statistically significant differences among groups (P = .009, a = .01). The Mann-Whitney U test showed significant differences between Admira Flow group and Tetric Flow (P = .011, a = .05), Nexus 2 (P = .001, a = .01), and X-Flow (P = .004, a = .01), respectively. Conclusion: Within the limitations of this study, the extent of microleakage in the cervical area (enamel) of Class II cavities restored with Sonicsys approx ceramic inserts depends on the material used for luting. The highest leakage occurred when Admira flow was used. Parole chiave: cervical microleakage, resin-based materials, Sonicsys ceramic inserts