OBJECTIVES: The purpose of this study was to investigate the effectiveness of multiple decontamination procedures for salivary contamination after curing of a universal adhesive on dentin bond strength according to its etch modes. MATERIALS AND METHODS: Forty-two extracted bovine incisors were trimmed by exposing the labial dentin surfaces and embedded in cylindrical molds. A universal adhesive (All-Bond Universal, Bisco) was used. The teeth were randomly divided into groups according to etch mode and decontamination procedure. The adhesive was applied according to the manufacturer's instructions for a given etch mode. With the exception of the control groups, the cured adhesive was contaminated with saliva for 20 sec. In the self-etch group, the teeth were divided into three groups: control, decontamination with rinsing and drying, and decontamination with rinsing, drying, and adhesive. In the etch-and-rinse group, the teeth were divided into four groups: control, decontamination with rinsing and drying, decontamination with rinsing, drying, and adhesive, and decontamination with rinsing, drying, re-etching, and reapplication of adhesive. A composite resin (Filtek Z350XT, 3M ESPE) was used for filling and was cured on the treated surfaces. Shear bond strength was measured, and failure modes were evaluated. The data were subjected to one-way analysis of variation and Tukey's HSD test. RESULTS: The etch-and-rinse subgroup that was decontaminated by rinse, drying, re-etching, and reapplication of adhesive showed a significantly higher bond strength. CONCLUSIONS: When salivary contamination occurs after curing of the universal adhesive, additional etching improves the bond strength to dentin.