Purpose: To assess the influence of bonding system and restorative material on the marginal integrity and pull-off forces of monolithic all-ceramic crowns bonded to titanium base (Ti-base) abutments.

Materials and methods: A total of 108 Ti-bases were sandblasted and divided into 9 experimental groups (n = 12 each) according to the combination of crown material (polymer-infiltrated ceramic network [PI], lithium disilicate [LD], and zirconia [ZI]) and bonding system (Multilink Hybrid Abutment [mh], PANAVIA V5 [pv], RelyX Ultimate [ru]), with their respective primers. After bonding the crowns to the Ti-base abutments, the restorations were screw-retained on implants and thermomechanically aged (1,200,000 cycles, 49 N, 1.67 Hz; 5°C to 55°C). Marginal integrity and bonding failure were evaluated under a light microscope, and pull-off forces (N) were calculated. Chi-square tests for marginal integrity, as well as one-way and two-way ANOVA for the pull-off forces, were applied (α = .05).

Results: PI presented higher marginal integrity than LD (P = .023). The bonding system pv revealed higher marginal integrity than mh (P = .005) and ru (P = .029). Differences in pull-off forces were found between restorative material and resin cement (P < .001), with the highest values for ZI + ru (598 ± 192 N), PI + pv (545 ± 114 N), LD + mh (532 ± 116 N), and PI + ru (528 ± 81 N). Specimens with marginal integrity revealed higher pull-off forces than those with alteration (P = .006). Specimens presenting bonding failures (micromovements) showed lower pull-off forces than those without bonding failures (P < .001).

Conclusion: The tested CAD/CAM materials show favorable bonding performances with different bonding systems; nevertheless, a specific bonding system has to be recommended for each restorative material.