This in vitro study aimed at evaluating the effect of glass-fiber post (GFP) surface treatment with a silane-containing universal adhesive and/or a silane coupling agent on push-out bond strength to dual cure resin cement/dentin (Experiment 1) and to resin composite (Experiment 2). Methods: For Experiment 1, sixty human premolar root canals were prepared to receive GFP cementation. Then an adhesive system was applied, according to the manufacturer's instructions. The GFPs (Reforpost no. 2, Angelus) were randomly divided into 2 groups, according to application or non-application of a silane coupling agent for 60 seconds (Prosil, FGM). Each group was further divided into 3 groups, according to the adhesive system to be applied to the GFP surface (n = 10): 1. SBU - silane-containing universal adhesive (Adper Single Bond Universal Adhesive, 3M ESPE); 2. ASB2 – etch-and-rinse adhesive system not containing silane in its composition (Adper Single Bond 2, 3M ESPE); 3. CG (control group) – no treatment. Cementation of GFPs to intraradicular dentin was performed with dual-cure conventional resin cement (Rely X ARC/ 3M ESPE). After 48 h, the roots were cross-sectioned at three different depths, resulting in serial slices corresponding to the cervical, middle and apical root thirds. For Experiment 2, surface-treated GFPs (n=10) were centered in a plastic matrix in which the resin composite (Filtek Z250 XT, 3M ESPE) was incrementally inserted and light-cured. Slices were obtained, as in Experiment 1. Push-out bond strength testing was performed on a universal machine (0.5 mm/min crosshead speed). Results: In Experiment 1, three-way ANOVA (α = 0.05) indicated that the bond strength of GFP to resin cement/intraradicular dentin was not affected by use or nonuse of a silane coupling agent, by application or non-application of an adhesive layer or by the root canal depth (p > 0.05). For Experiment 2, two-way ANOVA and Tukey's test showed that when the GFP received silane pretreatment, bond strength to resin composite was not affected by the application of an adhesive layer to the post. However, when the post was not pretreated with silane, the bond strength was higher for the universal adhesive (p = 0.021). The bond strength to post/resin cement/dentin was unaffected by post silanization or by the application of an adhesive system layer on the post. Bond strength of the post to resin composite was unaffected by adhesive system application, if the post was previously treated with a silane agent. In the absent of post silanization, a universal adhesive containing silane achieved higher bond strength of the post to resin composite.