The purpose of this study was to measure the effect rebonding, with and without etching restoration margins, had on the microleakage of class 5 composite restorations. Eighty-two class 5 composite restorations were placed in extracted human molars with one margin on etched enamel and one below the cementoenamel junction. After finishing with disks, the teeth were divided into seven margin treatment groups: 1) restorations with no rebonding (control), 2) restoration margins rebonded with Universal Bond 3 primer, 3) restoration margins etched with 37% phosphoric acid and rebonded with Universal Bond 3 primer, 4) restoration margins rebonded with Universal Bond 3 adhesive, 5) restoration margins etched and rebonded with Universal Bond 3 adhesive, 6) restoration margins rebonded with Fortify, and 7) restoration margins etched and rebonded with Fortify. After storage in a chlorine-based disinfectant for 1 week, the restorations were thermocycled and then soaked in dye. After sectioning, microleakage was assessed by measuring the extent of dye penetration along the restoration-tooth interface. At the enamel margins there was no statistical difference in microleakage rebonded with or without etching. On the gingival margins there was a statistically significant reduction in microleakage for the nonetched Universal Bond 3 adhesive rebonded group compared to the nonrebonded control.