Crop planting is high consumer of water and energy resources and a major contributor to anthropogenic greenhouse gas emissions (GHG). Expanding demand for food increases dependence on these resources, which in turn leads to water scarcity and increased GHG emissions. To address these challenges, optimization models were developed based on resource nexus. This study incorporates water and carbon footprints into an optimization framework for agricultural planting structures in arid regions, integrating Life Cycle Assessment (LCA) with Nondominated Sorting Genetic Algorithm II (NSGA-II), aiming to achieve optimal water conservation, greenhouse gas emission reduction, and improved grain production. The findings reveal that variations in agricultural inputs and management practices result in a higher carbon footprint per unit of production for cotton than for maize and wheat by 38.8 % and 56.4 %, respectively. The agricultural water footprint in arid regions is predominantly composed of blue water, with spatial heterogeneity driven by regional climatic differences. Carbon emission intensity, agricultural economic development levels, and water resource utilization efficiency were identified as the major drivers of the increase in water and carbon footprints in Xinjiang over the past 20 years. By reallocating sown areas of major crops to establish an optimal planting configuration, reducing arid agricultural carbon emissions by 11.4 % and curbing blue water usage by 12.6 %. The results provide a scientific basis and feasible solutions for achieving efficient resource use and environmental protection in arid farming systems while ensuring food security.