Purpose: To examine the effect of different adhesive luting systems on the shear bond strength of all-ceramic restorations to dentin surfaces. Materials and Methods: Forty-eight all-ceramic disks (2 x 3 mm; IPS e.max Press) were fabricated. Forty-eight noncarious extracted human molars were divided into 4 groups. In groups 1 to 4, IPS e.max Press disks were luted with Variolink 2/Excite DSC (etch-and-rinse), Clearfil Esthetic Cement/Clearfil Protect Bond (antibacterial and self-etching), Multilink/Multilink Primer (self-etching), or Multilink Sprint (self-adhesive). All specimens were subjected to 5000 thermocycles (5°C to 55°C, 30-s dwell time). Shear bond strengths were tested using a universal testing machine until failure. The analysis of the fractured dentin surfaces was performed using an optical microscope at 10X and 1000X magnification; the images were analyzed with an image analyzer. Data were analyzed with one-way ANOVA and Tukey's test at a significance level of p 0.05. Results: Mean shear bond strength data of the groups in MPa were: Variolink 2/Excite DSC: 25.89 ± 3.71; Clearfil Esthetic Cement/Clearfil Protect Bond: 17.21 ± 2.71; Multilink/Multilink Primer: 11.6 ± 3.51; Multilink Sprint: 10.4 ± 3.15. According to the one-way ANOVA, there were significant differences in shear bond strength (p 0.001). According to Tukey's test, statistically significant differences were observed in shear bond strength between groups 1 and 2 (p 0.001), groups 1 and 3 (p 0.001), groups 1 and 4 (p 0.001), groups 2 and 3 (p = 0.003), and groups 2 and 4 (p 0.001). Conclusion: The etch-and-rinse dentin bonding system produced higher bond strengths of all-ceramics to dentin surfaces than did the self-etching bonding systems and self-adhesive luting system. Parole chiave: shear bond, self-adhesive, dentin bonding, all-ceramic