Purpose: To determine the microleakage of etch and rinse adhesives. Materials and Methods: Standardized Class II cavities were cut in 40 human molars with one proximal box limited within enamel and one proximal box extending into dentin. Teeth were assigned randomly to 5 groups (n = 8) and restored with incrementally placed composite restorations. Five combinations were tested: G1 = XP Bond + Ceram X Mono, G2 = Syntac Classic + Tetric EvoCeram, G3 = Scotchbond 1 XT + Z250, G4 = P&B NT + Ceram X mono, G5 = Optibond Solo Plus + CeramX Mono. After finishing and polishing, teeth were stored for 48 h in water at 37°C before being subjected to artificial aging by thermal stress (5/55°C, 2000x, 30 s) and mechanical loading (50 N, 50,000x). Teeth were isolated with nail varnish and immersed in 5% methylene blue for 1 h. After sectioning, specimens were evaluated for leakage (ordinal scale: 0 to 4) at enamel and dentin margins under a stereomicroscope. Results were analyzed using the Kruskal-Wallis H-test and Mann-Whitney U-test (p 0.05). Results: Statistical analysis showed significant differences among the groups in both enamel and dentin. Mean ranks (H-test) were: enamel: G2 (64.44) G1 (66.69) G4 (74.88) G3 (98.25) and G5 (98.25); dentin: G3 (65.53) G1 (74.42) G4 (81.09) G2 (81.84) G5 (99.61). Conclusion: Microleakage of XP Bond is at the same level as or even better than other etch-and-rinse adhesives. Parole chiave: Class II restorations, dentin adhesives, composite, microleakage