Purpose: To investigate the efficacy of different ceramic surface cleaning methods after saliva contamination on the resin bond strength to lithium disilicate ceramics. Materials and Methods: 300 e.max CAD blocks (Ivoclar Vivadent) were polished with 600-grit silicon carbide paper and divided into five groups with or without human saliva contamination and according to the surface treatment performed (n = 10); control: no pretreatment; MP: Monobond Plus; PA+MP: 37% phosphoric acid (PA) followed by MP; HF+MP: 5% hydrofluoric acid (HF) followed by MP; MEP: Monobond Etch & Prime. The specimens were bonded with one of three resin cements: Variolink Esthetic DC (VE), Multilink Automix (MA) and Speed CEM (SC). After 24-h water storage, tensile bond strength (TBS) was measured. The ceramic surfaces after pretreatment were analyzed using x-ray photoelectron spectroscopy (XPS). Results: XPS analysis showed similar elemental distributions between saliva contamination vs no saliva in PA, HF, and MEP. The TBSs were significantly influenced by surface treatments (p 0.05). HF+MP and MEP showed statistically non-significantly different bond strengths to saliva-contaminated HF+MP and MEP, but were different from MP and saliva-contaminated MP. The TBSs after 24 h were significantly higher in HF+MP and MEP groups with VE. HF+MP and MEP did not show statistically significant differences among any groups with or without saliva contamination. Conclusion: Surface treatments with PA or HF followed by silane or by MEP alone were effective in removing saliva contamination and enhancing the resin bond strength. Keywords: lithium disilicate ceramics, saliva-contaminated, tensile bond strength