This study compared the ability of a variety of light sources and exposure modes to polymerize a dual-cured resin composite through ceramic discs of different thicknesses by depth of cure and Vickers microhardness (VHN). Ceramic specimens (360) (Empress 2 [Ivoclar Vivadent], color 300, diameter 4 mm, height 1 or 2 mm) were prepared and inserted into steel molds according to ISO 4049, after which a dual-cured composite resin luting material (Variolink 11 [Ivoclar Vivadent]) with and without self-curing catalyst was placed. The light curing units used were either a conventional halogen curing unit (Elipar TriLight [3M/ESPE] for 40 seconds), a high-power halogen curing unit (Astralis 10 [Ivoclar Vivadent] for 20 seconds), a plasma arc curing unit (Aurys [Degre K] for 10 seconds or 20 seconds) or different light emitting diode (LED) curing units (Elipar FreeLight I [3M/ESPE] for 40 seconds, Elipar FreeLight II [3M/ESPE] for 20 seconds, LuxOmax [Akeda] for 40 seconds, e-Light [GC] for 12 seconds or 40 seconds). Depth of cure under the ceramic discs was assessed according to ISO 4049, and VHN at 0.5 and 1.0 min distance from the ceramic disc bottom was determined (ISO 6507-1). Medians and the 25(th) and 75(th) percentiles were determined for each group (n=10), and statistical analysis was performed using the Mann-Whitney-U-test (p <= 0.05). The results showed that increasing ceramic disc thickness had a negative effect on the curing depth and hardness of all light curing units, with hardness decreasing dramatically under the 2-mm thick discs using LuxOmax, e-Light (12 seconds) or Aurys (10 seconds or 20 seconds). The use of a self-curing catalyst is recommended over the light-curable portion only, because it produced an equivalent or greater hardness and depth of cure with all light polymerization modes.