This study evaluated the Knoop hardness of a dual-cured resin cement (Rely-X ARC) activated solely by chemical reaction (control group) or by chemical / physical mode, light-cured through a 1.5 mm thick ceramic (HeraCeram) or composite (Artglass) disc. Light curing was carried out using conventional halogen light (XL2500) for 40 s (QTH); light emitting diodes (Ultrablue Is) for 40 s (LED); and Xenon plasma arc (Apollo 95E) for 3 s (PAC). Bovine incisors had their buccal face flattened and hybridized. On this surface a rubber mold (5 mm in diameter and 1 mm in height) was bulk filled with the resin cement. A polyester strip was seated for direct light curing or through the discs of veneering materials. After dry storage in the dark (24 h 37°C), the samples (n = 5) were sectioned for hardness (KHN) measurements, taken in a microhardness tester (50 gF load 15 s). The data were statistically analyzed by ANOVA and Tukey's test (α = 0.05). The cement presented higher Knoop hardness values with Artglass for QTH and LED, compared to HeraCeram. The control group and the PAC/Artglass group showed lower hardness values compared to the groups light-cured with QTH and LED. PAC/HeraCeram resulted in the worst combination for cement hardness values.