The efficacy of newly developed adhesive restorative materials is commonly described by in vitro test results such as bond strength and microleakage data. Due to the inconsistency often found between in vitro testing and in vivo performance, the purpose of the present investigation was to restore teeth scheduled for extraction in vivo with an experimental restorative bonding system and to evaluate the microleakage of the restorations by common laboratory techniques following extraction after 3 months of clinical service. Two restorations were placed in each of seven anterior teeth, one in the middle of the labial enamel portion and the other labially extending over the cementoenamel junction (CEJ). Six of the seven fillings completely surrounded by enamel showed no microleakage. However, two of the restorations crossing the CEJ revealed microleakage originating from the cervical margins. In spite of the small number of fillings assessed, the experimental enamel-dentin adhesive compound tested is preliminarily considered a promising new system.