This study evaluated the effect of surface conditioning methods on the microtensile bond strength (mu TBS) of a restorative composite to indirect restorative materials. Blocks (5 x 5 x 4 mm(3)) (N = 72) of (a) Zirconia (In-Ceram Zirconia, Vita) (ZR), (b) lithium disilicate glass ceramic (IPS Empress II, Ivoclar Vivadent) (LD), (c) Indirect resin composite (Gradia, GC) (GR) were fabricated (n = 24 per group) and divided randomly into three groups: 1-Control: no conditioning, 2-Silane coupling agent, 3-Hydrofluoric acid (9.5%) (HF)+silane. Each block was duplicated in resin composite. The adhesion surfaces were conditioned with airborne-particle abrasion (110 mu m Al2O3 particles). Half of the conditioned blocks received no bonding and the other half one coat of bonding (ED Primer II, Kuraray). Each conditioned block was bonded to a composite block with a resin luting agent (Panavia F2.0, Kuraray). The blocks were sectioned into 1 mm(2) microsticks and tested for microtensile bond strength (mu TBS) (0.5 mm/min) in a mu TBS testing machine. Failure types were evaluated under stereomicroscope and Scanning Electron Microscope (SEM). Data were analyzed using three-way ANOVA, Bonferroni corrected and independent sample t-tests (p < 0.05). Significant effect of the bonding (p < 0.001) and surface conditioning (p < 0.001) were observed in all groups. The highest mean bond strength values were obtained in the bonded, HF etched and silanized groups of ZR, LD and GR (12.4 +/- 2.9, 28.1 +/- 1.5 and 27.2 +/- 2 MPa, respectively). HF acid + silane increased the repair bond values in all materials. Majority of the failure types were adhesive for ZR group, whereas HF + silane conditioned LD and GR groups presented predominantly cohesive failures in the cement.