OBJECTIVE To determine and compare the shear bond strength of four single-bottle adhesives to enamel and dentin contaminated with human saliva. METHODS Prime & Bond 2.1, One Step, Tenure Quik and Syntac Single Component were evaluated. Ninety-six sound molars were divided into three subgroups for each of the four adhesives. Small flat areas on the buccal and lingual surfaces were ground in enamel. Small flat areas on the mesial and distal surfaces were ground on the middle dentin of the same teeth. Each adhesive was tested under: uncontaminated conditions (Group 1), contamination of the bonding surface with fresh whole saliva for 20 seconds and removing the excess with cotton pellets (Group 2), and contamination of the bonding surface with fresh whole saliva for 20 seconds and rinsing the saliva for 20 seconds before adhesive application (Group 3). After adhesive application, a plastic mold was filled with composite, applied to the bonding surface and light-cured. The teeth were thermocycled (1,000 cycles) and sheared using an Instron testing machine. The results were analyzed using a three-way ANOVA. RESULTS Saliva contamination of enamel or dentin did not affect the shear bond strength of the adhesives tested except for Syntac SC with contaminated dried enamel. One Step displayed similar bond strengths to both enamel and dentin; P & B 2.1, Tenure Q and Syntac SC displayed higher bond strength to enamel than to dentin. P & B 2.1 showed significantly higher bond strengths to enamel than the other adhesives, but there was no significant difference for dentin bond strength under the uncontaminated condition.