The purpose of this study was to evaluate the capability and characteristics of different nanoleakage observation methods, including light microscope (LM), field-emission scanning electron microscope (FESEM), transmission electron microscope (TEM), and confocal laser scanning microscope (CLSM). Dentin specimens were bonded with either an etch-and-rinse adhesive (SBMP) or a self-etch adhesive (GB), and prepared for nanoleakge evaluation according to different observation methods. LM, FESEM and CLSM results demonstrated that the SBMP group showed more interfacial nanoleakage than the GB group (p<0.05); by contrast, no significant difference was found in TEM results (p>0.05), however, TEM illustrated concrete nanoleakage forms or patterns. The results suggested that different observation methods might exhibit distinct images and a certain degree of variations in nanoleakage statistical results. Researchers should carefully design and calculate the optimum assembly in combination with qualitative and quantitative approaches to obtain objective and accurate nanoleakage evaluation.