This study evaluated and compared microleakage values of in vivo and in vitro placed Class I amalgam restorations with or without three different lining materials and posterior composite restorations with two dentin bonding agents. For the in vivo group, 72 standardized Class I cavities were prepared on the occlusal surfaces of molars scheduled for extraction. The test groups (n = 12) were: amalgam without lining (A), amalgam with cavity varnish (A+C), amalgam with Clearfil SE Bond (A+CSE), amalgam with Clearfil 2V (A+C2V), composite with Clearfil SE Bond (C+CSE) and composite with Protect Bond (C+PB). The restored teeth were extracted after seven days. The same grouping, materials and techniques were used in 72 extracted molars for the in-vitro part of the study. The specimens were immersed in basic fuchsin for 24 hours and sectioned. Microleakage was examined and scored at 20x magnification. Statistical analyses were performed with the Kruskal-Wallis and Mann-Whitney U-tests with the 5% level of significance. Overall, the in vivo and in vitro test groups were not different from each other. No significant differences in microleakage values were observed between the unlined and lined amalgam groups (p > 0.05). However, since lined amalgam restorations did not reveal any marginal leakage, the application of an adhesive bonding material under the amalgam restorations can be considered. In general, cavity varnish was not as effective as adhesive bonding agents in preventing microleakage in amalgam restorations. Composite restorations demonstrated higher leakage values than amalgam restorations (p < 0.05), except for A+C (p > 0.05) in the in vivo group. There was no significant difference between the two composite groups for in vitro and in vivo conditions (p > 0.05).