RESUMEN
Aim: To evaluate the occlusal and internal marginal adaptation of inlay restorations made of different materials, using CAD-CAM. Methods: Preparations were made for MOD inlays of one-third intercuspal width and 4 mm depth in 30 third human molars. The teeth were restored using CAD-CAM materials (n=10) of nanoceramic resin (Lava Ultimate), polymer-infiltrated ceramic network (VITA ENAMIC), or lithium disilicate glass-ceramic (IPS e.max CAD). The specimens were cemented with dual resin cement and sectioned at the center of the restoration, after which the two halves were evaluated, and photographed The occlusal and internal discrepancy (µm) was determined at five points: cavosurface angle of the occlusal-facial wall (CA-O); center of the facial wall (FW); faciopulpal angle (FPA); center of the pulpal wall (PW); and center of the lingual wall (LW). The data were submitted to the Kruskal-Wallis and the Dunn tests (α=0.05). Results: No difference was observed among the materials regarding the occlusal discrepancy at the CA-O, FPA, or PW internal points. The e.max CAD measurement at FW showed larger internal discrepancy than that of Lava (p=0.02). The internal discrepancy at LW was greater for e.max CAD than VITA ENAMIC (p=0.02). Conclusion: Lithium disilicate glass-ceramic presented greater internal discrepancy in relation to the surrounding walls of the inlay preparations