Vegetation is the only component that can be directly managed to reduce the negative consequences of wildland fires. Flammability indexes provide information about plant flammability characteristics and are commonly used in wildland fuel management. However, previous flammability indexes were usually based on only two or three flammability components. We propose an advanced approach for leaf flammability index estimation that integrates all four flammability components (ignitability, combustibility, sustainability and consumability) using cluster and linear discriminant analyses. We measured time to ignition, ignition temperature, maximum flame height, combustion time, mass loss percentage and fuel moisture content of 15 plant species most affected by forest fires in Turkey (every two weeks, from May to September). Species were divided into different flammability classes and ranked according to their flammability index (FI) values. Classifications and FI rankings showed seasonal variations for some of the species. During May and June, fuel moisture contents explained similar to 50-60% of variation in leaf flammability but this relationship steadily decreased towards September (13%) when weather conditions became drier. The proposed approach for FI estimation offers a more detailed flammability assessment and comparison between plants. It can be used to identify less flammable plants for fuel breaks and landscaping in the wildland-urban interface.