Objective: To evaluate the effect of different surface pretreatments on the marginal microleakage of Vitremer restorations. Study Design: Class V cavities were prepared on the labial and lingual surfaces of extracted human third molar teeth. Cavities were randomly distributed into six groups according to surface treatment, as follows: Group NC (negative control): no surface treatment; Group VP (positive control): Vitremer Primer; Group PA: 32% phosphoric acid; Group PAA: 20% polyactylic acid; Group PLP: self-etch adhesive (Prompt L Pop); Group PB: etch&rinse adhesive (Prime&Bond NT). All cavities were restored with Vitremer. Teeth were thermocycled, stained with 0.5% basic fuchsin dye and sectioned. Microleakage values were quantitatively assessed by linear measurement of dye penetration using image-analyzing software. Differences between occlusal and gingival microleakage values within groups were evaluated using paired t-tests, and differences among groups were analysed using one-way analysis of variance (ANOVA) and Tukey tests. Results: There were statistically significant differences between the occlusal and gingival microleakage values for all groups (p<0.05). Micro leakage differed significantly among surface pretreatment groups (p<0.001). The negative control and PLP groups showed similar microleakage values, but were significantly higher than other groups for both margins. Although there were no statistically significant differences between positive control and PA, PAA, PB groups, microleakage values of positive control group were smaller than all other experimental groups except for PB group. Conclusions: Vitremer restorations require surface pretreatment to prevent excessive microleakage. Pretreatment with etch&rinse adhesives and Vitremer Primer may reduce microleakage of Vitremer restorations.