Recent studies have argued that changes in fire regimes in the 21st century are posing a major threat to global biodiversity. In this scenario, incorporating species’ physiological, ecological, and evolutionary traits with their local fire exposure might facilitate accurate identification of species most at risk from fire. Here, we developed a framework for identifying the animal species most vulnerable to extinction from fire-induced stress in the Brazilian savanna. The proposed framework addresses vulnerability from two components: (1) exposure, which refers to the frequency, extent, and magnitude to which a system or species experiences fire, and (2) sensitivity, which reflects how much species are affected by fire. Sensitivity is based on biological, physiological, and behavioral traits that can influence animals’ mortality “during” and “after” fire. We generated a Fire Vulnerability Index (FVI) that can be used to group species into four categories, ranging from extremely vulnerable (highly sensible species in highly exposed areas), to least vulnerable (low-sensitivity species in less exposed areas). We highlight the urgent need to broaden fire vulnerability assessment methods and introduce a new approach considering biological traits that contribute significantly to a species’ sensitivity alongside regional/local fire exposure.