OBJECTIVE The aim of this in vitro study was to evaluate the influence of different light curing units and modes on microleakage of flowable composite resins. METHODS Eighty Class V cavities were prepared in buccal and lingual surfaces of 40 extracted human premolars with cervical wall located in dentin and the occlusal wall in enamel. These teeth were randomly assigned into two groups (n=20) and restored with different flowable composites; Group I: Esthet-X Flow, Group II: Grandio Flow. Each group was randomly divided into four subgroups; while the samples of the first subgroup were polymerized with conventional Halogen light, the rest of them were polymerized with different curing modes of Light Emitting Diode (LED). The second subgroup was polymerized with fast-curing; the third subgroup with pulse-curing and those of the fourth subgroup with step-curing modes of LED. After the samples were thermocycled and immersed in dye, they were longitudinally sectioned. Dye penetration was assessed under a stereomicroscope. Data were analyzed by Kruskal-Wallis and Mann-Whitney U tests. RESULTS None of the restorations showed leakage on enamel margins. On dentin margins no significant differences were observed between flowable composite resins polymerized with halogen light (P>.05). While step curing mode of LED presented significant differences between the resins, the difference was insignificant when fast-curing and pulse-curing mode of LED were used. No statistically significant differences were observed between curing units for Esthet-X Flow samples. For Grandio Flow samples, only step-curing mode of LED caused statistically higher leakage scores than halogen and other curing modes of LED (P<.05). CONCLUSIONS The effect of curing units' type and curing mode on flowable composite resin leakage might be material-dependent.