Due to its vast territory, the climatic conditions and socioeconomic development of different regions in China are closely related to geographical location, and their impact on city-level building carbon emissions (BCEs) shows significant spatial differences. Ignoring this difference will hinder the building sector from achieving carbon peak and carbon neutrality goals. This study integrates climate and economic factors into a unified analysis framework from a geographical perspective, uses the global regression model ordinary least squares (OLS), local regression models geographically weighted regression (GWR) and multiscale geographically weighted regression (MGWR) for comparative verification, and explores the gradient impact of China's climate and economic geographical factors on city-level BCEs. Furthermore, BCE reduction policy recommendations for different regions are proposed. The model comparison results show that the effect obtained when using the global regression model to analyze influencing factors related to geographical location may be biased. The local MGWR results show that the impacts of climate geographical factors heating degree days (HDD) and cooling degree days (CDD) on city-level BCEs have obvious north‒south gradient characteristics, while the impacts of economic geographical factors population (P), tertiary industry GDP (GDP3), urbanization rate (UR) and floor area (FR) on city-level BCEs show obvious gradient decreasing characteristics from northeast to southwest. These gradient characteristics cannot be ignored when promoting building carbon emission reduction work. This study expands and deepens the driving mechanism of BCEs and provides a reference for formulating building carbon emission reduction policies based on local conditions.