OBJECTIVE: To assess four different chemical surface conditioning methods for ceramic material before bracket bonding, and their impact on shear bond strength and surface integrity at debonding. METHODS: Four experimental groups (n = 13) were set up according to the ceramic conditioning method: G1 = 37% phosphoric acid etching followed by silane application; G2 = 37% liquid phosphoric acid etching, no rinsing, followed by silane application; G3 = 10% hydrofluoric acid etching alone; and G4 = 10% hydrofluoric acid etching followed by silane application. After surface conditioning, metal brackets were bonded to porcelain by means of the Transbond XP system (3M Unitek). Samples were submitted to shear bond strength tests in a universal testing machine and the surfaces were later assessed with a microscope under 8 X magnification. ANOVA/Tukey tests were performed to establish the difference between groups (α= 5%). RESULTS: The highest shear bond strength values were found in groups G3 and G4 (22.01 ± 2.15 MPa and 22.83 ± 3.32 Mpa, respectively), followed by G1 (16.42 ± 3.61 MPa) and G2 (9.29 ± 1.95 MPa). As regards surface evaluation after bracket debonding, the use of liquid phosphoric acid followed by silane application (G2) produced the least damage to porcelain. When hydrofluoric acid and silane were applied, the risk of ceramic fracture increased. CONCLUSIONS: Acceptable levels of bond strength for clinical use were reached by all methods tested; however, liquid phosphoric acid etching followed by silane application (G2) resulted in the least damage to the ceramic surface.