This in vitro study compared the depth of cure obtained with six quartz tungsten halogen and light-emitting diode curing units at different exposure times and light tip-resin composite distances. Resin composite specimens (Tetric Ceram, A3; diameter 4 mm, height 6 mm) were exposed from 0-, 3-, and 6-mm distance. The curing units (200-700 mW/cm2) were used for standard (20 and 40 s), pulse-delay mode (initial exposure of 3 s at 200 mW/cm2, followed by a resting period of 3 min and a final exposure of 10 or 30 s at 600 mW/cm2), or soft-start curing (40 s; exponential ramping). Curing depth was determined by measurement of Wallace hardness for each half millimeter starting at 0.5 mm from the top surface. For each specimen, a mean H(W) value was calculated from the H(W) values determined at the depths of 2.0 mm and less (0.5, 1.0, 1.5, and 2.0 mm, respectively). The depth of cure for each specimen was found by determining the greatest depth before an H(W) value exceeding the minimal H(W) value by 25% occurred. For all curing units, an increase in exposure time led to significantly higher depth of cure. Increasing the light tip-resin composite distance significantly reduced the depth of cure. With a light tip-resin composite distance of 6 mm, median values of depth of cure varied between 2.0 and 3.5 mm following a 20-s (or 3+10 s) exposure and between 3.0 and 4.5 mm following a 40-s (or 3+30 s) exposure. The composite situated above the depth of cure value cured equally well with all curing units. At both exposure times, Luxomax resulted in the significantly lowest depth of cure, and Astralis 7 yielded significantly higher depth. At both exposure times, a significant linear correlation was found between the determined power densities of the curing units and the pooled depth of cure values obtained. It seems that for the resin composite tested, the recommended exposure time of 40 s per 2-mm increment may be reduced to 20 s, or that increments may be increased from 2 to 3.5 mm. It may be that the absolute values of depth of cure found are material specific, but we believe that the relationships found between curing units, between exposure times, and between light guide distances are universal.