Understanding the multi-scale characteristics and multi-dimensional interaction mechanisms of Resource and Environmental Carrying Capacity (RECC) is critical for achieving sustainable development goals (SDGs). However, existing studies lack the integration of SDGs into the RECC evaluation framework and pay limited attention to interaction mechanisms and synergistic trade-offs among RECC dimensions in arid regions. We developed a comprehensive framework that integrated the SDGs into a "pressure-dual support" coupled multi-model approach for multi-scale RECC assessment. Moreover, the coupling relationships, trade-offs, and dynamic impact effects among RECC dimensions were explored using the coupling coordination degree (CCD) model, correlation analysis, and a panel vector autoregression model. The results showed that Xinjiang's RECC improved significantly across all scales from 2000 to 2022, with distinct south-to-north differences in RECC levels and significant spatial autocorrelation and clustering effects at smaller scales. The CCD among RECC dimensions exhibited a notable upward trend, with regional differences observed in CCD improvements across different scales. While strong synergies among dimensions were observed across Xinjiang, these relationships shifted toward trade-offs as spatial scale decreased, with significant regional variations in interactions. Furthermore, we found that social carrying capacity was a key driving factor, promoting economic carrying capacity while constraining resource carrying capacity and environmental carrying capacity. These results support our recommendation that policymakers should adopt differentiated regional management strategies to enhance RECC, promote coordination across its multiple dimensions, and strengthen the role of social carrying capacity to advance green and sustainable development.