The effectiveness and the failure mechanism of fire barriers in residential upholstered furniture (RUF) were investigated by full-scale flaming tests on single upholstered chair mock-ups. Six fire barriers were tested. The fire barriers were screened for (1) the presence of elements that are typically used in fire retardants and, (2) the presence of targeted fire retardants. For each fire barrier, triplicate flammability tests were run on chair mock-ups where polyurethane foam and polyester fiber fill were used as the padding materials, and each chair component was fully wrapped with the fire barrier and a polypropylene cover fabric. The ignition source was an 18 kW square propane burner, impinging on the top surface of the seat cushion for 80 s. Results showed all six fire barriers reduced the peak heat release rate (up to 64 %) and delayed its occurrence (up to 19 min) as compared to the control chair mock-ups. The heat release rate remained at a relatively low plateau level until liquid products (generated by either melting or pyrolysis of the padding material) percolated trough the fire barrier at the bottom of the seat cushion and ignited, while the fire barrier was still intact. The peak heat release rate occurred shortly after. These observations indicated that the failure mechanism of the fire barrier was related to the ignition of the percolating liquid under the chair.