<p>In the context of energy justice, energy poverty remains a pressing concern in emerged China’s dual-carbon transition, with substantial regional disparities persisting in energy access and utilization. This study develops a comprehensive indicator system for energy poverty in China. Using K-means clustering, the 30 provinces are classified into four categories. Kernel density estimation, Dagum’s Gini coefficient with decomposition, and the Geodetector model are then employed to examine the dynamic evolution, regional disparities, and driving factors of energy poverty. The analysis yields four main findings. First, China’s energy poverty index exhibits an oscillating upward trajectory, with regional disparities forming a gradient pattern in the sequence of Category II, Category IV, Category I, and Category III regions. Second, spatial dynamics reveal significant heterogeneity, as regional polarization and divergence coexist with a tendency toward convergence. Third, the overall Gini coefficient displays a fluctuating increase, highlighting persistent imbalances that are mainly driven by inter-regional differences, while intra-regional disparities and hypervariable density contribute less. Finally, economic development, urbanization, innovation capacity, openness, and industrial upgrading are identified as principal contributing factors, and their interactions exert stronger effects on energy poverty than individual factors alone. The findings not only deepen the understanding of energy poverty and energy justice in developing countries but also provide policy insights for narrowing inter-regional disparities, promoting inclusive urbanization and industrial upgrading, and strengthening institutional mechanisms that ensure equitable and sustainable access to energy.</p>