•We investigate the relationship between burned area and fire-weather for the entire globe, focusing on the fire season the grid-box scale.•A large proportion of total global land area exhibited a significant relationship between fire weather index (FWI) and burned area.•The most sensitive areas to fire weather fluctuations are those with a FWI below 17, like tropical moist and temperate broadleaf forests and also boreal forests and taigas.•We calculated near-future FWI scenarios (2026–2045) using an ensemble of state-of-the-art Earth System Models from the CMIP5 database.•Projected FWI anomalies for the most sensitive biomes are either negative or close to zero, although positive increments in some regions may result in significant disruptions in fire regimes in boreal and Amazon rain forest among others. Fire is an integral Earth system process, playing an important role in the distribution of terrestrial ecosystems and affecting the carbon cycle at the global scale. Fire activity is controlled by a number of biophysical factors, including climate, whose relevance varies across regions and landscapes. In light of the ongoing climate change, understanding the fire–climate relationships is an issue of current interest in order to identify the most vulnerable regions. Building upon recent global observations of burned areas and climate, we investigate the sensitivity of fire activity to fire–weather across the world's major biomes. We identify the biomes susceptible to inter-annual fire–weather fluctuations, unveiling a non-linear relationship with a saturation threshold past which the area burned can be considered insensitive to increasing fire-weather. Our results depict an unambiguous spatial pattern that identifies the world regions where short-term climate fluctuations are unlikely to produce any significant effect on current fire activity, and those most sensitive to fire-weather changes. In particular, the boreal forests and extensive areas of tropical and subtropical moist broadleaf forests (excluding Africa) as well as sizeable areas of temperate broadleaf forests are identified as highly sensitive. We then present near-future fire-weather scenarios (period 2026–2045) using a state-of-the-art ensemble or Earth System Models (ESMs) from the CMIP5 database, considering a moderate and a high emission scenario (RCPs 4.5 and 8.5). The projected anomalies in fire-weather for the vulnerable temperate biomes are small in magnitude and their direction is either negative or just slightly positive, although significant differences in the projected probability density functions suggest that disruptions of fire regimes may occur locally. Other sensitive ecosystems, like the rain forests of the Amazon basin may experience a significant increment in fire-weather that may result in severe impacts on fire regimes as a direct consequence of climate change in the next decades.