The North Atlantic Basin (NAB) is seeing a significant increase in the frequency and intensity of tropical cyclones since the 1980s, with record-breaking seasons such as 2017 and 2020. However, little is known about how coastal ecosystems, particularly mangroves in the Gulf of Mexico and the Caribbean, are responding to these new climate normals at regional and subregional scales. Wind speed, rainfall, pre-cyclone forest structure, and hydro-geomorphology are known to influence mangrove damage and recovery following cyclones in the NAB. However, these studies have focused on site-specific responses and individual cyclonic events. Here, we analyze 25 years (1996-2020) of mangrove vulnerability (damage after a cyclone) and short-term resilience (recovery after damage) for the entire NAB and its subregions, using multi-annual, remote sensing-derived databases. We applied machine learning to characterize the influence of 22 potential drivers that include previously researched variables and new ones such as human development and long-term climate trends. The characteristics of the cyclones mainly drive vulnerability at the regional level, while resilience is largely driven by site-specific conditions. These include long-term climate conditions, such as air temperature and drought trends, pre-cyclone habitat conditions, such as canopy cover and height and soil organic carbon stock, and human interventions on the land. Rates and drivers of mangrove vulnerability and resilience vary across subregions in the NAB, and hotspots for restoration and conservation actions are highlighted within subregions. The impacts of increasing cyclone activity need to be framed in the context of climate change compound effects and heavy human influences in the region. There is an urgent need to value the restoration and conservation of mangroves as fundamental Nature-based Solutions against cyclone impacts in the NAB.