Objective: The aim of this study was to assess the bond strength of adhesive systems to dentin contaminated by temporary cements with or without eugenol. Method: Flat dentin surfaces were obtained from twenty-four human third molars. With exception of the control group (n=8), the surfaces were covered with Interim Restorative Material (Caulk Dentsplay, Milford, DE, USA) or Cavit (3M ESPE, St. Paul, MN, USA) and kept in an oven at 37°C for seven days. After removing the cements, the adhesive systems Adper Single Bond (3M ESPE, St. Paul, MN, USA) or Clearfil SE Bond (Kuraray Co. Ltd., Osaka, Japan) were applied in accordance with the manufacturers' recommendations, and then the crowns were constructed in of resin composite. The teeth were sectioned into specimens with a cross-sectional bond area of 0.81mm2, which were submitted to microtensile testing in a mechanical test machine at an actuator speed of 0.5mm/min. The data were analyzed by t- and ANOVA tests, complemented by Tukey tests (α=0.05). Results: For Adper Single Bond (3M ESPE, St. Paul, MN, USA), bond strength did not differ statistically (p>0.05) for all the experimental conditions. For Clearfil SE Bond (Kuraray Co. Ltd., Osaka, Japan), only the Interim Restorative Material (Caulk Dentsplay, Milford, DE, USA) Group showed significantly lower bond strength (30.1 ± 13.8 MPa) in comparison with the other groups; control (38.9 ± 13.5 MPa) and Cavit (3M ESPE, St. Paul, MN, USA) (42.1 ± 11.0 MPa), which showed no significant difference between them. Conclusion: It was concluded that the previous covering of dentin with temporary cement containing eugenol had a deleterious effect on the adhesive performance of the self-etching system only.