This study assessed the finishing and polishing of 3 ceramic materials: Vitadur Alpha, IPS Empress 2 and AllCeram. Surface modification techniques were selected to simulate dental practice. Forty-five specimens of each ceramic were divided into 5 groups of 9 specimens, which were finished using the following procedures: Group 1--glaze; Group 2--glaze, grinding and subsequent polishing with the Eve system; Group 3--glaze, grinding and subsequent polished with the Identoflex system; Group 4--glaze followed by polishing with Identoflex; Group 5--glaze, grinding and subsequent polishing with Shofu kit. Two roughness-measuring instruments were used: a stylus profilometer and an atomic force microscope (AFM). The 135 specimens were evaluated quantitatively with respect to Ra (average roughness) and Ry (maximum roughness height), and the results were examined statistically by ANOVA and Tukey's test, with a significance level of 0.05. The roughness parameter (Ra), measured by the profilometer, and AFM showed that some of the commercial intraoral polishing kits tested achieved a finish equal in smoothness to the glazed surface. According to Pearson's test, no correlation was found between the parameter Ry, measured with the profilometer, and AFM. The profilometer results for Ry demonstrated no significant differences between the final polished surfaces and the initial glazed ones. On the other hand, the Ry values obtained by AFM indicated the tested polishing kits incapability of producing smoothness comparable to the glazed surfaces.