Purpose: To evaluate the ability of the provisional filling material Cavit-W alone or in combination with different restorative materials to prevent bacterial leakage through simulated access cavities in a resin buildup material. Materials and Methods: LuxaCore resin cylinders were subdivided into 4 experimental groups (n = 30), plus a positive (n = 5) and a negative (n = 30) control group. One bore hole was drilled through each cylinder, except those in the negative control group (G1). The holes were filled with Cavit-W (G2), Cavit-W and Ketac-Molar (glassionomer cement, G3), Cavit-W and LuxaCore bonded with LuxaBond (G4), Cavit-W and LuxaCore (G5), or left empty (G6). Specimens were mounted in a two-chamber leakage setup. The upper chamber was inoculated with E. faecalis. An enterococci-selective broth was used in the lower chamber. Leakage was assessed for 60 days and compared using Fisher's exact test (α 0.05) corrected for multiple testing. Results: Bacteria penetrated specimens in the positive control group within 24 h. All specimens in the negative control group resisted bacterial leakage for 60 days. Twenty-seven specimens in G2, 26 in G3, and 16 specimens in G5 showed bacterial leakage by the end of the experiment. G4 prevented bacterial penetration completely. The statistical comparison revealed significant differences between G4 and all other experimental groups. Conclusion: Under the current conditions, Cavit-W alone or combined with a glass-ionomer cement did not prevent bacterial leakage through a resin buildup material for two months. In contrast, covering Cavit-W with a bonded resin material resulted in a bacteria-tight seal for two months. Schlagwörter: core buildup composite, temporary restoration, bacteria, leakage