Methods: TC specimen was 0.8 mm thick porcelain veneer, color ETC2 (corresponding to A3) (n=5) of IPS-Empress ceramics in combination with “natural die material” in color ND4 (represents average dentin) (n=10) and neutral cement color MV0 (Variolink), all from Ivoclar/Vivadent, Shaan, Lichterstein. CM specimens were 0.8 mm thick ET1 (translucent glossy), ETC0 (translucent + shade), ETC2 (translucent + shade A1, A2), EO1 (average opacity) and EO3 (very high opacity) in combination with 4 discolored dentins ND5, ND6, ND7 and 2 cements (low and high value). Color was measured with SpectroShade (MHT, Italy) in the imitated clinical situation. One-way ANOVA for dL*, da*, db* and deltaE was performed to evaluate the influence of background and cement. DeltaE<2.6 was utilized to define color match between TC and CM specimens. Results: The statistical analysis showed that the background (dentine color) had significant contribution to the overall color (p<0.05). The color shifts (a* of b*) depended more on the dentine- than on the veneer color, try-in cement did not significantly alter the final color and opaque ceramics showed less color variation due to their masking ability. Color match was the closest using opaque ceramics when dentine was slightly discolored. It was possible to achieve TC by discolored dentine when lighter veneer was used. Conclusions: The try-in cement had no influence on color prediction whereas the use of different ceramic translucencies and lightness could predictable mask discolored dentin in order to achieve pre-selected color match.