Background: When partial coverage restorations (veneers, inlays, onlays...) must be cemented to dentin, bond strength may not reach the same predictable values as to enamel. The purpose of this study was: 1. To compare, with a shear bond test, the bond strength to dentin of a total-etch and a self-etching bonding agent. 2. To determine whether creating microretention improves the bond strength to dentin. Materials and methods: Two bonding agents were assayed, Optibond FL® (Kerr), two-bottle adhesive requiring acid etching, and Clearfil SE Bond® (Kuraray), two-bottle self-etching adhesive. The vestibular, lingual, distal and mesial surfaces of ten molars (n=10) were ground to remove all enamel and 40 ceramic samples were cemented with Variolink II® (Ivoclar Vivadent). Half the molar surfaces were treated to create round microretention (pits) to determine whether these could influence bond strength to dentin. The 40 molar surfaces were divided into four groups (n=10): Optibond FL (O); Clearfil SE (C); Optibond FL + microretention (OM); Clearfil SE + micro retention (CM). A shear bond test was performed and the bond failures provoked examined under an optical microscope. Results: O=35.27±8.02 MPa; C=36.23±11.23 MPa; OM=28.61±6.27 MPa; CM=27.01±7.57 MPa. No statistically significant differences were found between the adhesives. Optibond FL showed less statistical dispersion than Clearfil SE. The presence of microretentions reduced bond strength values regardless of the adhesive used. Conclusions: 1. Clearfil SE self-etching adhesive and Optibond FL acid-etch showed adequate bond strengths and can be recommended for bonding ceramic restorations to dentin. 2. The creation of round microretention pits compromises these adhesives’ bond strength to dentin.