<p>Western China, as an energy resource base, depends heavily on its resources for economic development, and Shaanxi Province—an energy transmission hub in this region—typifies efforts to achieve regional carbon peaking and neutrality goals with corresponding spatiotemporal evolutionary patterns in energy consumption carbon emission efficiency (ECCEE). This study simulates carbon emissions in Shaanxi prefecture-level cities from 2008 to 2023 using DMSP-OLS-like nighttime light data. ECCEE is measured using a slack-based measurement (SBM) undesirable model and employed kernel density estimation with Markov chains to reveal its spatiotemporal evolution. The spatial Durbin model (SDM) and its effect decomposition to examine the spatial interactive impacts of different factors. Based on an “energy supply–demand technology” three-dimensional framework, dynamic qualitative comparative analysis (QCA) is applied to explore multiple efficiency enhancement pathways. The research findings revealed a U-shaped trajectory (an initial decline followed by recovery) in Shaanxi Province’s ECCEE, reflecting an adaptation process of the energy system transition. Spatially, the province exhibited a pattern of central regions leading and northern and southern areas following, with an agglomeration tendency in high-efficiency cities. The estimation results of the SDM indicate that ECCEE presents significant positive spatial dependence. Three high-efficiency driving pathways are indentified and further refined them into two models: a “demand technology-driven” model and a “technology driven” model. These empirical findings highlight that targeted technological energy-saving innovation, carbon-reduction process upgrades, clean energy structure optimization, and industrial restructuring are vital for enhancing ECCEE in resource-dependent regions.</p>