Effect of desensitizers on the bond strength of resin cements to dentin was evaluated. Intact premolars (N=90) were embedded in polymethyl methacrylate; dentin surfaces were exposed, and they were randomly divided into two main groups of cements (Duolink (D), Variolink II (V); n=45 per group) and then into three desensitizer subgroups (n=15 per subgroup). Teeth in controls (C) were treated according to cements' adhesion protocols; the other two groups received either fluoride[Aqua-Prep F (F)] or triclosan-based [Seal&Protect (T)] desensitizers. Ceramic disks (Empress 2) were adhered; specimens were thermocycled (x5,000 cycles, 5-55 +/- 1 degrees C, dwell time 30 s) and subjected to shear bond strength test (MPa +/- SD) in a universal testing machine (crosshead speed 1 mm/min). Failure types were classified using scanning electron microscope. For V, application of both desensitizers (29.6 +/- 7.8 and 22.8 +/- 2.8 for F and T, respectively) did not present significantly different results than that of the VC (21.2 +/- 2.3; p>0.05, one-way ANOVA). In D, F (20.6 +/- 2.4) showed significantly higher results (p<0.05) than those in T (16.1 +/- 3.9) and DC group (15.2 +/- 2.3). V showed significantly higher results than D (p<0.05, Bonferroni). F and T did not negatively affect the bond strength results with D and V. Adhesive failures were more frequent with both T (84%) and F (66%) in D; cohesive failures in the cement (88%) were more commonly observed with F in V. Both F and T desensitizers can be safely used prior to final cementation but F in combination with V seems to be more reliable, considering both the bond strength and the failure types.