OBJECTIVE This investigation was carried out to evaluate the long-term effects of fluoride-releasing adhesive resins on structural changes in standardized fluid-filled gaps simulating microleakage between the materials and the tooth surface in vitro. METHODS Three commercially available fluoride-releasing resin adhesives (One-Up Bond F, OptiBond Solo, and Reactmer Bond) were used in this study. Cured disks of resin adhesive were placed over flat human tooth surfaces (enamel and dentin), separated by a standardized 40microm interfacial gap and stored in distilled water for 24h (control group) or 1000 days (experimental group). After 1000 days of water storage, the resins were detached from the teeth and the opposing surfaces were examined by scanning electron microscopy (SEM). In addition, chemical structural analysis was performed by laser-Raman spectroscopy. RESULTS The SEM microphotographs showed numerous crystal types on the enamel, dentin, and resin surfaces after 1000 days of water storage for OptiBond Solo and Reactmer Bond. However, there was no crystal formation in the control specimens and the aged specimens of One-Up Bond F. Raman analysis showed several peaks (463, 618, and 990cm(-1)) from the crystals of OptiBond Solo that were not identified in the enamel, dentin, or cured resin. CONCLUSIONS In conclusion, two of the three tested fluoride-release resin adhesives (OptiBond Solo and Reactmer Bond) have the ability to induce crystal growth within gaps between the adhesive and teeth in long-term water storage. These results suggest that the two adhesive resins have self-reparative ability with regard to bond leakage.