To evaluate the influence of occlusal load cycling on cervical microleakage of proximal slot restorations located in dentin, using two self-etching and two one-bottle dentin adhesive systems. 240 proximal slot cavities were prepared in 120 bovine teeth and divided into two groups, one with load cycling and one without. The groups were then subdivided into four subgroups according to the adhesive system used (Experimental EXL 547 Self-etching 3M, Clearfil SE Bond, Single Bond, and Optibond Solo Plus) and restored following the manufacturers' instructions. The teeth were then submitted to mechanical load cycling with a force of 80 N and a frequency of 5 Hz, simultaneously over both restorations of each tooth, for a total of 50,000 cycles per specimen. All specimens were subsequently immersed in a 2% methylene blue solution (pH 7.0), and sectioned to examine the extent of dye penetration under a stereomicroscope (40X). There was no statistically significant difference (p = 0.00002) between the loaded and unloaded teeth. However, a statistically significant difference was observed between the adhesive systems used. The experimental self-etching EXL 547 presented the lowest mean microleakage, but was only statistically significantly different from the Single Bond loaded and unloaded groups and the Clearfil SE Bond unloaded group. The application of 50,000 loading cycles did not affect the microleakage of the two self-etching and the two one-bottle adhesive systems evaluated. In vitro mechanical load cycling is an important factor to consider when evaluating the performance of adhesive systems under simulated masticatory conditions.