The purpose of this study was to determine the effects of contamination with smoker's and non-smoker's saliva on the bond strength of resin composite to superficial dentin using different adhesive systems. The interfacial structure between the resin and dentin was evaluated for each treatment using environmental scanning electron microscopy (ESEM). Freshly extracted human molars were ground with 600-grit SiC paper to expose the superficial dentin. Adhesives [One-Up-Bond-F-Plus (OUFP) and Adper-Prompt-L-Pop (APLP)] and resin composite (TPHSpectrum) were bonded to the dentin (n = 8/group, 180 total specimens) under five surface conditions: control (adhesive applied following manufacturers' instructions); saliva, then 5-s air dry, then adhesive; adhesive, saliva, 5-s air dry; adhesive, saliva, 5-s water rinse, 5-s air dry (ASW group); and adhesive, saliva, 5-s water rinse, 5-s air dry, reapply adhesive (ASWA group). After storage in water at 37 degrees C for 24 h, the specimens were debonded under tension at a speed of 0.5 mm/min. ESEM photomicrographs of the dentin/adhesive interfaces were taken. Mean bond strength ranged from 8.1 to 24.1 MPa. Fisher's protected least significant difference (P = 0.05) intervals for critical adhesive, saliva, and surface condition differences were 1.3, 1.3, and 2.1 MPa, respectively. There were no significant differences in bond strength to dentin between contamination by smoker's and nonsmoker's saliva, but bond strengths were significantly different between adhesive systems, with OUFP twice as strong as APLP under almost all conditions. After adhesive application and contamination with either smoker's or nonsmoker's saliva followed by washing and reapplication of the adhesive (ASWA group), the bond strength of both adhesive systems was the same as that of the control group.