<p>Amid rapid urbanization and growing socio-environmental risks, understanding urban resilience dynamics is crucial for sustainable city governance. This study proposes a data-driven framework to evaluate the socio-spatial resilience of Chengdu’s central area from 2000 to 2020. Integrating the entropy weighting method and set pair analysis, it constructs an eight-dimensional evaluation indicator system. Results reveal an overall increase in comprehensive resilience, though marked by pronounced heterogeneity across social areas and dimensions. Middle-class and working-class social areas achieved multi-dimensional improvements driven by industrial upgrading and infrastructure renewal, while migrant commercial and tenant social areas remained vulnerable due to weak institutional and social foundations. Meanwhile, low social capital resilience in relevant social areas has resulted from large-scale urban renewal that dismantled traditional community networks and forced the relocation of disadvantaged populations. This trajectory highlights a distinct transformation in the dominant drivers of comprehensive resilience amid the transition from a planned economy to a market economy: evolving from institution-based protection, to market-driven adaptive adjustment, and eventually to a renewed phase of institutionally supported resilience enhancement. Moreover, the evolution of resilience aligns with the adaptive cycle model: rapid growth during the exploitation–conservation (r–K) phase was followed by a release (Ω) of systemic rigidity in vulnerable areas, triggering social capital depletion, and eventually a reorganization (α) phase characterized by community rebuilding and institutional adaptation. These findings provide empirical support for differentiated and resilience-oriented urban governance strategies in rapidly urbanizing inland cities.</p>