Context: A significant forest decline has been noticed these last years in Europe. Managers need tools to better anticipate these massive events. Objectives: We evaluated the efficiency of easily available data about environmental conditions and stand characteristics to determine different levels of vulnerability. Methods: We combined remote sensing images, photo-interpretation, and digital models describing environmental conditions within a modelling approach to achieve spatial vulnerability assessment of the stands. We focused on silver fir and Norway spruce stands in the Vosges mountains (8,900 km², northeastern France), where severe symptoms of decline are visible. Results: Silver fir were predicted highly vulnerable on 7% of their area versus 33% for Norway spruce. Using an independent dataset, we observed ten-times (silver fir) and two-times (Norway spruce) higher mortality rates in the units with a high level of vulnerability than in the others. About half of the model deviance was directly or indirectly explained by variables related to water stress (soils displaying low water availability, having suffered severe drying events these last years). Furthermore, the stands acclimatised to drought conditions were more resilient. Stand characteristics also influenced dieback spread, suggesting that an evolution of silvicultural practices toward mixed stands with broadleaved species and uneven-aged trees can contribute to better adapt to future climate conditions. Conclusions: Vulnerability maps based on easily available geographic information describing climate, soil, and topography can efficiently discriminate canopy mortality patterns over broad areas, and can be useful tools for managers to mitigate the effects of climate change on forests.