<p>Road traffic carbon dioxide (CO₂) has a fundamental role in global warming. Accurately estimating and understanding the spatio-temporal patterns of urban road traffic CO₂ emissions plays a fundamental role in developing targeted reduction strategies. However, few studies have estimated CO₂ emissions at the urban road scale with fine spatio-temporal resolution. Therefore, this study adopted a bottom-up method to estimate urban road traffic CO₂ emissions using Global Positioning System (GPS) trajectory data. Urban road traffic CO₂ emissions from individual vehicles are estimated using a vehicle trajectory-driven CO₂ emission model The aggregated results are mapped within Traffic Analysis Zones (TAZ) to create CO₂ emission distribution maps with minute-level temporal and road-level spatial resolution. The Multiscale Geographically Weighted Regression (MGWR) model is employed to analyze the impact of various elements of the built environment on urban road transport CO₂ emissions. Experimental results indicate that road traffic CO₂ emissions in Hangzhou have spatio-temporal heterogeneity. Road traffic CO₂ emission hotspots are concentrated along main roads such as Shixiang Road, the City Ring Expressway, and Shiqiao Road. Further analysis indicates that population density, main road density, availability of bus stops, and length of bike lanes exert a significant influence on urban road transport CO₂ emissions in Hangzhou. These findings enhance our recognition of the combined effects of the various elements of the built environment on urban road transport CO₂ emissions. This study introduces a method for estimating CO₂ emissions at the street level using vehicle trajectory information. It provides high spatio-temporal resolution CO₂ emission distribution maps to support carbon emissions reduction strategies in urban transportation.</p>