This study evaluated the effect of polymerization mode and time and thermal and mechanical loading cycling (TMC) on microleakage in composite resin restorations. One hundred and eighty cavities were prepared and randomly divided according to the light curing time (20, 40, or 60 s), modes (quartz-tungsten-halogen (QTH)-420 mW/cm(2), LED 2 (2nd degree generation)-1,100 mW/cm(2), or LED 3 (3rd degree generation)-700 mW/cm(2)), and TMC. Following standard restorative procedures, the samples were prepared for analysis in an absorbance spectrophotometer. All results were statistically analyzed using the three-way ANOVA and Tukey test (p ≤ 0.05). The results revealed that the groups QTH and LED 3 submitted to TMC showed higher microleakage than those that were not submitted to TMC. Only for LED 3, 60 s showed higher microleakage than 20 s. For LED 2 and QTH, there were no differences between the times. QTH showed lower microleakage means than LED 2, when photoactivated for 20 s, without TMC. When photoactivated for 60 s, QTH showed lower microleakage means than LED 3, for the groups with or without TMC. It was concluded that TMC, the increase in polymerization time, and the irradiance were factors that may increase the marginal microleakage of class II cavities.