The objectives of this study include comparing the cement thickness and microleakage of Class II ceramic inlays built with three ceramic systems and verifying whether there was a correlation between those two variables. The ceramic systems used include: 1) Heat-pressed (IPS-Empress); 2) CAD-CAM (CEREC 2) and 3) Sintered (Colorlogic). Standardized MOD Class II inlay cavities with one proximal box extending below and the other extending above the cement-enamel junction (CEJ) were prepared in 30 extracted human molars and randomly assigned to three groups. The ceramic inlays were constructed according to manufacturer's instructions and cemented using a dual-cure resin cement (Variolink II). All teeth were mechanically cycled (100,000 cycles, 78N) and thermocycled (700 cycles, 5 degrees C-55 degrees C). After immersion in silver nitrate, the inlays were sectioned mesial-distally and evaluated with an optical microscope (40x). The cement thickness obtained by the Colorlogic system (enamel: 113 +/- 25 microm; dentin: 118 +/- 23 microm) was significantly higher than that obtained by CEREC (enamel: 78 +/- 14 microm; dentin: 87 +/- 13 microm) and Empress (enamel: 65 +/- 15 microm; dentin: 89 +/- 14 microm). Regarding dye penetration, there was no statistical difference among the three ceramic systems in enamel. At the dentin margins, the Colorlogic system resulted in a significantly higher penetration depth compared to CEREC and Empress, which had similar average values. No correlation was found between cement thickness and microleakage either in enamel or dentin for any of the ceramic systems.