To evaluate the effects of high-intensity light curing units (light-emitting diode [LED] and plasma arc curing [PAC]) on the microleakage of flexible spiral wire retainers (FSWRs) at the composite/enamel and composite/wire interfaces. Forty-five human mandibular incisor teeth were separated into three groups of 15 teeth. Multistranded PentaOne wire of .0215 inch diameter was bonded to enamel and was cured with three different light curing units: a quartz-tungsten-halogen (QTH) unit and two high-intensity units (ie, LED and PAC). A conventional halogen light served as the control. Samples were sealed with nail varnish, stained with 0.5% basic fuchsine, and sectioned. Transverse sections were evaluated under a stereomicroscope and were scored for microleakage for the composite/enamel and composite/wire interfaces. Statistical analysis was performed by Kruskal-Wallis and Mann-Whitney U-tests with Bonferroni correction. Little or no microleakage was detected at the composite/enamel interface of the FSWR cured with three different light sources. However, at the composite/wire interface, statistically significant differences were found between the QTH (mean, 1.10 +/- 1.05 mm) and high-intensity curing units. The PAC resulted in the greatest amount of microleakage (mean, 2.63 +/- 1.49 mm), whereas no statistically significant difference was noted between the PAC and the LED (mean, 2.35 +/- 1.28 mm). High-intensity light curing units show statistically significant microleakage at the composite/wire interface and therefore may not be safe for use in bonding FSWRs.