The purpose of this study was to test the hypothesis that mechanical polishing methods of ceramic surfaces allow similar superficial roughness to that of glazed surfaces. Twenty-five Vitadur Alpha ceramic discs (5 mm x 2 mm) were prepared according to the manufacturer's specifications. All specimens were glazed and randomly assigned to 5 groups (n=5), according to finishing and polishing protocols: G1: glazed (control); G2: diamond bur finishing; G3: G2 + silicon rubber tip polishing; G4: G3 + felt disc/diamond polishing paste; G5: G3 + felt disc impregnated with fine-particle diamond paste. Next, surface roughness means (Ra - microm) were calculated. Qualitative analysis was made by scanning electron microscopy. Surface roughness data were submitted to ANOVA and Tukey's test at 5% significance level. G1 and G4 were statistically similar (p>0.05). G2 presented the highest roughness means (p<0.05) followed by groups G3, G5, G4 and G1 in a decreasing order. The hypothesis was partially confirmed as only the mechanical polishing (G4) produced similar superficial roughness to that of surface glazing, although finishing and polishing are technically critical procedures.