Objective: To evaluate the influence of different cement combinations on coronal microleakage in restored endodontically treated teeth using dye penetration. Method and Materials: Human, noncarious single-rooted extracted premolars (n = 60) were divided into four experimental groups (each n = 15). After endodontic treatment, different combinations of cements were used to lute prefabricated posts and complete crown restorations: zinc phosphate cement applied on posts and cast crowns (Z) or on zinc phosphate cement posts and resin cement applied on cast crowns (ZR); resin cement applied on posts and zinc phosphate cement applied on cast crowns (RZ); and resin cement applied on posts and cast crowns (R). After artificial aging through thermal cycling (5°C to 55°C) for 2,000 cycles at 38 seconds for each cycle and 15 seconds of dwell time, specimens were immersed for 72 hours in basic fuchsin at 37°C. A buccolingual section was made through the vertical axis of specimens. A Toolmaker's microscope (Mitutoyo) was used to measure (µm) dye penetration. The Kruskal-Wallis nonparametric test was used to determine intergroup difference. A nonparametric Mann-Whitney test compared each group regarding its maximal linear penetration depths on the mesial and distal aspects of each specimen (α = 0.05). Results: Dye staining was evident to some degree in all specimens. Among groups Z, ZR, and RZ, no significant difference was shown in dye-penetration depths (mean penetration scores 1,518 to 1,807 µm). However, dyepenetration depth was significantly lower in group R compared to the other groups (mean penetration score 1,073 µm) (P .05). Conclusions: Under study conditions, the cement combination offering the best coronal sealing was the one using only resin cement for both posts and crown restorations. Palabras clave: crowns, luting agents, posts, restorations