To assess the degree of nanoleakage of Class V preparations restored with three bonding agents using wet and dry bonding. Standardized Class V cavities were prepared in 66 human molars with cervical margins located in dentin. The composite was placed using one of three different dentin bonding agents with a moist or dry bonding technique (n = 11 per group). The teeth were stored in a 1% rhodamin-B solution for 24 h at 20°C, then rinsed with water, embedded in methacrylate, and sectioned parallel to the long axis of the tooth, separating the restorations into two parts. A confocal laser scanning microscope (CLSM) was used to visualize the layer 10 µm under the prepared surface of the section. The lengths of the dye-penetrated pathways were measured, representing the amount of nanoleakage. Differences in nanoleakage between the tested materials were proven to be statistically significant. The effect of drying had a significant influence on both the acetone-based and the ethanol-based material, but not on the water-based bonding resin. Dentin drying may have an influence on nanoleakage, depending upon the nature of the individual bonding agent.