Historically, Class II restorations using resin composite with margins below the dentino-enamel junction (dej) have fared poorly in investigations, exhibiting significant dye penetration along the marginal interfaces. In all reports located by the authors the gingival margins of sub-CEJ restorations have been finished, even though in actual practice this is not feasible in the vast majority of cases. The purpose of this study was to examine the effect of finishing on gingival margins located below the cemento-enamel junction of Class II resin composite restorations. Class II resin composite restorations with margins below the dej were placed in 40 extracted human molars using a dentin-enamel adhesive, a flowable resin composite, and a universal microhybrid resin composite restorative material. In all groups the adhesive was light activated for ten seconds and the restorative resin composite was light activated for ten seconds. The flowable increments were 1 mm in thickness and the restorative resin composite increments were 2 mm in thickness. In groups 1, 3, and 4 the flowable was light activated for ten seconds while in group 2 the flowable was light activated for 40 seconds. In groups 1, 2, and 4 the sub-CEJ margins were finished with finishing discs while in group 3 the sub-CEJ margins were unfinished. In group 4 the dentin adhesive was reapplied following finishing procedures. All groups were immersed in dye for 24 hours. The two groups in which the margins were finished exhibited varying degrees of dye penetration. There was no dye penetration in the unfinished and the resealed groups. It was concluded that the finishing procedure itself causes damage to the resin-dentin interface, which allows dye penetration to occur. This could potentially explain why resin adhesive materials have fared so poorly in Class II in vitro investigations, which is not the common clinical experience.