<p>Mangrove forests offer nature-based solutions for climate change mitigation by storing atmospheric CO₂, while also supporting biodiversity, livelihoods, and coastal protection. Yet, they face increasing threats from deforestation, salinity, and extreme weather events. Here, we assessed and explored the resilience of Sundarbans mangrove forests and associated drivers using satellite-derived vegetation indices, biodiversity, and environmental data, within a structural equation modeling framework. We found 1–8 major perturbations at 250-meter spatial resolution, with the lowest resilience to disturbance or stress in the central and southeastern zones of the Sundarbans. Approximately 610 to 990 km² (~10-15% of the total Sundarbans area) of mangrove forests exhibited declining resilience. The functional composition of maximum canopy height (MCH) was the strongest driver of mangrove forest resilience (β= 0.61), followed by specific leaf area (SLA; β= 0.56) and precipitation (β= 0.44). Structural diversity (β= 0.22), though weakly associated, mediated the positive association of species richness with resilience. Perturbation frequency had a significant negative direct (β= −0.29) and total (β= −0.33) association on resilience, whereas temperature exhibited only a total negative association (β= −0.20). Based on these findings, we hold that for enhancing mangrove forest resilience against disturbances and stressors, (re-)establishment and conservation efforts should focus on maintaining site-specific dominant species with tall canopy height and specific leaf area —supplemented by a few complementary species— to increase structural diversity and facilitate recovery rates.</p>