This study compared the microleakage of light-cured and auto-set glass ionomer liners used in Class V composite laminated glass ionomer restorations by determining the amount of microleakage at the gingival cementum/dentin margins. Standardized nonundercut V-shaped Class V cavities with gingival margins below the cementoenamel junction were prepared on the mesial and distal surfaces of 40 molars, establishing a total of 80 cavities, which were randomly divided into four groups. Each was lined with glass ionomers: group 1, Ketac-Bond (ESPE-Premier), which served as the control; group 2, XR-Ionomer (Kerr); group 3, XR-Ionomer with polyacrylic acid (PAA) pretreatment (Kerr); and group 4, Vitrabond (3M). Specimens were thermocycled for 300 cycles in 0.5% aqueous solution of basic fuchsin between 4 and 55 C with a 1-minute dwell time, and individually embedded in an epoxy resin. Statistical analysis indicated no differences among groups using the light-cured glass ionomer (groups 2 to 4), and they showed significantly less leakage than the control (group 1) at P less than 0.00001). Removal of the smear layer using 10% polyacrylic acid solution did not influence microleakage in restorations with light-cured glass ionomer liners.