Aim:This study evaluated the effect of filler existence in self-etch adhesive resin on the marginal leakage of a class V restoration.Materials and Methods:Class V cavities were prepared and restored with a resin composite on the buccal surfaces of 48 premolars lined with unfilled or filled adhesives (n = 24). After thermo cycling, teeth in each group were divided to two subgroups (n = 12), specimens of the first subgroup were incubated for 24 h in distilled water at 37°C, and for the second group three months in the same condition. Specimens were placed in 50% silver nitrate for 24 h at 37°C, and then were cut buccolingually 1 mm thick. Dye penetration was measured using a stereomicroscope and scaled from 0 to 5 in a blind method. SEM images were made to evaluate the dentin-adhesive interfaces. Collected data were analyzed using the nonparametric Kruskal-Wallis and Mann-Whitney U-tests at a significant level of P<0.05.Results:There was no significant difference between microleakage of filled and unfilled adhesive at 24 h and 3 months (P<0.05). There was a significant difference in cervical microleakage between 24 h and 3 months, which was independ on filler load of the adhesive (P<0.001). In contrast, there was no significant difference in occlusal microleakage between 24 h and 3 months and the cervical microleakage was significantly higher than occlusal microleakage after 3 months. SEM images reveald that unfilled adhesive infiltrate slightly better than filled adhesive.Conclusion:The application of filler particles in a self etch adhesive system had no influence on marginal leakage at both the enamel and dentin margins. While the unfilled adhesive infiltrate better than the filled adhesive, its long term performance is not promising.