To comprehensively capture the spatio-temporal effects of urbanization on CO₂ emissions, we develop an innovative hybrid spatial model (SLX-GTWR) that incorporates spatial spillover effects and spatio-temporal heterogeneity. Using nighttime light data and a new dataset, we measure urbanization in terms of scale and intensity and estimate monthly fossil CO₂ emissions at a spatial resolution of 0.1° for 268 Chinese cities. The empirical results show that: (1) The direct effect, spatial spillover effect, and total effect of urbanization on CO₂ emissions are spatio-temporal heterogeneity. (2) In terms of temporal heterogeneity, the spatial spillover effects of urbanization on CO₂ emissions is negative and shows a trend of first increase and then decrease. (3) In terms of spatial heterogeneity, positive spatial spillovers occurred in cities around three major clusters, while the spillovers of the clusters were negative. Overall, the most cities in China has reached a stage where urbanization is conducive to carbon reduction. Our detailed results inform the development of targeted low-carbon urban development policies.