Methods and Materials: Disk specimens 2 mm in diameter and 2.5 mm thick were prepared from 2 composites (Vit-l-escence, shade A3, Ultradent; Herculite XRV, shade A2, Kerr) in the following manner. An extracted permanent molar tooth was prepared to receive a Class II restoration with proximal slot only. The tooth was sectioned horizontally on a plane above the gingival fl oor level such that the remaining depth of the proximal box was 4 mm. A Toffl emire matrix band and retainer were secured around the tooth. Composite specimens were placed below the tooth to coincide with the location of the slot opening at the bottom of the gingival fl oor. The specimens were subjected to light polymerization with various combinations of curing cycle and light unit: 1 of 3 continuous curing cycles (20 seconds, 40 seconds or 60 seconds) and 1 of 2 LED units (Utralume-5, Ultradent; IQ Smartlite, Dentsply) or a control quartz–tungsten–halogen (QTH) unit (Optilux 501, Kerr). Specimens were stored at 37°C for 24 hours. A hardness tester was used to obtain 4 measurements of Knoop hardness number (KHN) for each surface (upper and lower) of each specimen. Relative hardness (RH) was calculated as the KHN of the lower surface divided by the KHN of the upper surface. Data were analyzed with analysis of variance (ANOVA) and Tukey's test.

Results: ANOVA indicated signifi cant differences in mean RH among the groups (p < 0.001). RH increased with increasing curing time. For the 60-second cycle with Vit-l-escence composite, mean RH was 0.47, 0.25 and 0.39 for the Ultralume-5, IQ Smartlite and Optilux 501 curing units, respectively. For the 60-second cycle with Herculite XRV composite, mean RH was 0.71, 0.81 and 0.56 for the Ultralume-5, IQ Smartlite and Optilux 501 curing units, respectively.

Conclusions: In general, the 2 LED units performed as well as the QTH unit; however, brand of composite and curing cycle had signifi cant effects on RH values.