Objectives: The aim of the study was to evaluate the effect of different supstrate colors, luting agent colors and thicknesses, different thicknesses and translucency levels of core ceramics, and different thicknesses of veneering ceramics on the final color of a model representing lithium-disilicate restorations. Materials and methods: Cylindrical specimen of Natural Die material in the dimensions 11 x 5 mm were prepared in the colors ND1, ND3, ND5, ND7 and ND9 – 11 cylinders of each color. Disc specimen Ø 11 mm of IPS e.max Press LiS2 ceramics in 4 different thicknesses (0.5, 0.7, 1.0 and 1.2 mm) and 3 translucency levels (HT, LT and MO) were heat-pressed, and 3 thicknesses (0.3, 0.5, 0.8 mm) of IPS e.max Ceram veneering ceramics were prepared. That resulted in a total of 120 e.max Press specimen (4 thicknesses x 3 translucency levels x n = 10), 30 IPS e.max Ceram specimen (3 thicknesses x n =10), and 55 Natural Die specimen (5 colors x n = 11). One ND disc of each color served as control, while on the other 10 a layer of Variolink II resin cement in 5 different colors (transparent, white, white opaque, yellow and brown) and two thicknesses (80 or 150 μm) was applied. Different combinations of core discs with or without veneering ceramics were overlapped with the background discs with and without the cement layer to measure L*a*b* values using a spectrophotometer. Color differences ΔE* were calculated and data was statistically analyzed by means of STATISTICA 10.0. Analysis of variance (ANOVA) was applied. Results: A significant influence of supstrate color (p < 0,001), thickness (p < 0,001) and translucency level (p < 0,001) of LiS2 ceramics, color (p < 0,001) and thickness (p < 0,001) of luting agent, and the thickness of veneering ceramics (p < 0,001) on the color parameters was found. Despite the fact that the color and thickness of the luting agent had a statistically significant effect, only the white opaque cement caused clinically significant, i.e. visible color shifts. Veneering ceramics of varying thickness had a statistically significant impact on the color parameters of the model which simulated an all ceramic restoration, whereby, besides increasing lightness of the samples, a decrease of the clinical impact of underlying colors (supstrate and luting agent) was noticed. Conclusion: For optimum color-matching results the thickness of the restoration, supstrate, and conditionally luting agent color need to be taken into account.