The aim of the present study was to investigate the cytotoxicity of two "one bottle" adhesive systems after polymerization with a conventional halogen or a light emitting diode (LED) lamp. We hypothesized that different polymerization sources might enhance the intracellular production of reactive oxygen species (ROS), leading to reduced cell survival. Two "one bottle" adhesive systems (Optibond Solo and Scotchbond One) were cured with a commercial halogen (Optilux 500) and an LED source (Elipar Freelight, 3 M). The specimens were extracted for 24 h in complete cell culture medium or in phosphate-buffered saline (PBS). Endothelial cells (ECV 304) were exposed to the extracts for 24 h and survival rates were evaluated by the MTT assay. Then, ROS generation was monitored by the oxidation-sensitive fluorescent probe 2'7'-dichlorofluorescin diacetate (DCFH-DA). Extracts from all materials except for Optibond Solo polymerized with the halogen lamp were rated significantly cytotoxic. Scotchbond One cured with LED was the most toxic material, which reduced cell survival to about 23% compared with control cultures. Significantly higher amounts of ROS were produced in cell cultures treated with adhesives polymerized with the LED lamp compared with the materials cured with the commercial halogen light source. We demonstrated that the production of intracellular ROS by extracts of the adhesive systems depended on the light sources used for curing of the materials. These results suggested a possible link between ROS production and cytotoxic activity.