This study evaluated the microleakage of resin-modified glass ionomer, flowable compomer and flowable resin composite restorations on a Class V cavity of simulated advanced root caries under a flexural load cycling condition. Thirty-six non-carious human maxillary premolars were mounted in cylindrical acrylic resin molds. The cavities were prepared in the proximal root surface, from the middle of the buccal surface to the middle of the lingual surface, approximately 1 mm below the cemento-enamel junction, 2 mm axial width and 1.2 mm in depth. The teeth were randomly assigned to one of three groups with 12 teeth in each group: Group 1: Cavity conditioner and Fuji II LC (GC America), Group 2: Prime & Bond NT and Dyract Flow (Caulk-Dentsply), Group 3: Excite and Tetric flow (Ivoclar/Vivadent). Specimens were settled laterally on a fatigue-testing machine that was adjusted to deliver a force of 60N. The specimens were load cycled at 1 Hz for 5000 cycles, placed in a staining solution and sectioned to evaluate microleakage penetration. Results indicate that the coronal and gingival margins showed significant microleakage differences among the three restorations (p<0.05). At the coronal margin, there was no significant difference between Groups 2 and 3. At the gingival margin, there was no significant difference between Groups 1 and 2. It was concluded that the marginal sealing ability of a flowable resin composite under a flexural cycling load was better than in other selected materials and that flowable resin composite with dentin adhesive was a desirable alternative for root caries restorations extended to the proximal surface.