Materials and methods. Ninety class V cavities (1.5×2×3 mm) were prepared on the buccal surfaces of 90 bovine teeth. The teeth were randomly divided into 6 groups (n=15) and restored as follows: Groups 1 and 2 with Z100, groups 3 and 4 with Z250, and groups 5 and 6 with Nulite F composite resins. All the specimens were thermocycled. Groups 1, 3 and 5 were selected as control groups (without bleaching) and the experimental groups 2, 4 and 6 were bleached with 22% carbamide peroxide gel. All the samples were immersed in 2% basic fuchsin dye for 24 hours and then sectioned longitudinally. Dye penetration was evaluated under a stereomicroscope (×25), at both the gingival and incisal margins. Data were analyzed using Kruskal-Wallis, Mann-Whitney and Wilcoxon tests (alpha=0.05). 

Results. Statistical analyses revealed that bleaching gel increased microleakage only at gingival margins with Z250 (P=0.007). Moreover, the control groups showed a statistically significant difference in microleakage at their gingival margins. Nulite F had the maximum microleakage while Z250 showed the minimum (P=0.006). 

Conclusion. Microleakage of home-bleached restorations might be related to the type of composite resin used