OBJECTIVES:The purpose of this study was to compare nanoleakage associated with polished and fractured specimens, and evaluate the effects of thermocycling on nanoleakage. METHODS:The dentin bonding systems used in this study were PermaQuik and Prime and Bond NT (PBNT). Flat occlusal dentin surfaces and class V cavities from extracted human molars were bonded with one of the dentin bonding systems. With or without thermocycling, samples were immersed in a 50% w/v solution of silver nitrate for 24h, and exposed to photodeveloping solution for 8h. The specimens were cut or fractured, and mounted on stubs, carbon coated and observed in a Field Emission-SEM using backscattered electron mode. RESULTS:Polished and fractured specimens showed similar nanoleakage patterns for both dentin-bonding systems. Thermocycling with regimens of different temperatures and number of cycles did not result in different nanoleakage patterns and leakage lengths in controls and thermocycled groups. PBNT combined with non-rinse conditioner (NRC) produced no significant difference in nanoleakage length, but a somewhat different nanoleakage pattern, compared with PBNT and 34% phosphoric acid. SIGNIFICANCE:Thermocycling does not have an effect on nanoleakage of dentin bonding systems. The smearing effect from sectioning and polishing procedures does not affect the nanoleakage result.