Background: The aim of this study was to establish a new in vitro model for biofilm induced secondary caries studies using an oral biofilm reactor. Methods: An approximately 2 · 3 · 2 mm 3 sized dentino-enamel Class I cavity was prepared in the middle of a squareshaped specimen from the mid-labial portion of bovine incisors. The cavities were partially filled with either Clearfil AP-X with SE-Bond or Clearfil AP-X without any bond. Artificial biofilms were then formed on the resin composite filled surfaces using three species of oral bacteria in an oral biofilm reactor for 20 hours followed by 7-or 30-day incubation periods. Results: The lesions were clearly visible on fluorescence microscopy and by scanning electron microscopy in the enamel at the interface of resin restorations in all samples. The data from image analysis showed that the lesion size was largest in the No-bond samples with statistically significant differences (p < 0.05). Demineralization along the cavity wall extended deeper in No-bond compared to SE-Bond samples and penetration was significantly deeper in No-bond 30-day samples. Conclusions: A primary artificial secondary caries model was established using biofilms for in vitro studies and the significance of using a bonding system could also be verified.