Aim The aim of this study was to develop a test model to quantify the penetration of bacteria into dentinal tubules. Methodology The model consisted of two compartments separated by a bovine dentine specimen with a thickness of 1.5-3.1 mm. The root cementum was removed from the root surface and the specimens were oriented in the model with the pulpar side facing the inoculated chamber of the test model. One compartment contained the test organism and the other was tilled with sterile broth that was evaluated for growth of the test organism. The depth of bacterial penetration was measured in the dentine with or without a smear layer using both a histological and a quantitative recovering grinding technique, after 6 weeks of exposure to the microorganisms. Results E. faecalis penetrated dentine significantly deeper than A. israelii (P < 0.001), After removal of the smear layer with EDTA, E. faecnIisl penetrated significantly deeper than in dentine pretreated with saline: only (P < 0.01) or with a combination of saline and sodium hypochlorite (P < 0.01), Microorganisms were found in 89% of the cultured specimens and in 80% of the specimens that were evaluated with light microscopy. Total penetration through the dentine specimen and infection of the broth in the test compartment. of the model occurred in only two out of 72 specimens. Conclusion Collection and immediate culturing of infected dentine dust and counting colony forming units (CFU) allowed an overview of the number of bacteria per sample and was more sensitive than microscopy. Removal of the smear layer enhanced bacterial penetration.