<p>Rapid globalization and urbanization have accelerated land use change, challenging ecological sustainability and spatial balance in traditional heavy-industry cities. This study develops a transferable, multi-scale framework integrating land use suitability assessment with scenario-based spatial optimization, demonstrated in Taiyuan, China. We construct a natural and socio-economic indicator system using AHP–CRITIC hybrid weighting, the efficacy coefficient method, and a multi-factor overlay model at county and grid scales, and apply a geographical detector to identify dominant drivers and their interactions. Crucially, we employ the Future Land Use Simulation (FLUS) model to predict and optimize land use patterns under four 2029 scenarios—farmland protection, ecological conservation, green low‑carbon, and integrated development—parameterized by suitability surfaces and driver constraints and calibrated with 2011–2023 transitions. Results show agricultural and built-up land suitability peaks in the east, while ecological land suitability is higher in the north and west. Key drivers vary by land type—distance to water sources for agricultural land, slope for built-up land, and elevation for ecological land—with interaction effects exceeding single-factor influences. Land use transitions (2011–2023) reveal continuous grassland loss, expansion of other land categories, and small but directional centroid shifts. The integrated scenario achieves the most balanced outcome between ecological protection and urban growth. Policy recommendations include strengthening spatial control over built-up land, safeguarding cultivated land security, restoring sensitive ecosystems, and promoting green, resource-efficient urban transformation. The framework is replicable for improving land resource efficiency, guiding spatial restructuring, and advancing sustainable spatial governance in industrial cities globally.</p>