Questions: Sustaining ecosystem services for society in the face of global change is a major challenge. Resilience, an ecosystem's ability to absorb disturbance and return to its initial functioning, is critical for this. We assessed the resilience of mown subalpine grasslands following restoration management. Using a novel multi-indicator approach, we asked how different taxonomic and functional facets of biodiversity influence resilience mechanisms. We demonstrated this approach for a case study in subalpine grasslands, asking: how resilient are grasslands dominated by Patzkea paniculata and their agronomic services to changing mowing regimes? Location: Villar d'Arene, Hautes-Alpes, Provence-Alpes-Cote d'Azur, France. Boundary area of Ecrins National Park. Methods: A manipulative experiment tested the reciprocal effects of mowing cessation and mowing resumption for 10 years. We analysed floristic composition data following four steps. First, we used the Community Structure Integrity Index (CSII; Jaunatre et al., ) to obtain qualitative and quantitative summaries of taxonomic responses. The second step focused on taxonomic biodiversity using species richness, Simpson and Pielou indices, and responsive species identified in the first step. Third, we analysed functional diversity responses using functional groups and community weighted mean (CWM) of vegetative plant traits. Finally, we quantified ecosystem services impacts by estimating fodder quantity and quality using trait-based models. Results: The mowing manipulation demonstrated the high resilience of P. paniculata grasslands and revealed reversibility of transitions between mown and unmown states. By reducing the abundance of P. paniculata, the resumption of mowing restored forage quality. Supported by a complementary case study on post-disturbance re-sowing in hay meadows, this study demonstrates the resilience of mown subalpine grasslands to management change and explores underlying belowground mechanisms of vegetative regeneration and belowground reserves. Conclusion: Our novel multi-indicator approach provides multifaceted mechanistic understanding necessary to anticipate impacts of socio-ecological changes and to maintain the multiple benefits of mountain grasslands. Addressing the different facets of biodiversity from abundance data that are systematically collected in impact or monitoring assessment, this approach provides a common framework, widely applicable to different types of restoration or management interventions, across regions and biota.