Controlling carbon emissions is a vital step toward reaching Carbon Emissions Peak and Carbon Neutrality Targets, as well as advancing low-carbon development in urban areas. There exists a significant correlation between spatio-temporal changes of urban form and carbon emissions. Firstly, based on carbon emissions, land use and landscape pattern indices from 2000 to 2020, the spatio-temporal distribution maps of carbon emissions and urban form in Beijing were generated. Secondly, we built panel data models to estimate the impact of urban forms on carbon emissions. Finally, the optimal parameter Geodetector was employed to reveal the driving factors of urban form on carbon emissions. The results show that: (1) The spatial distribution of carbon emissions from 2000 to 2020 exhibited a pattern of high concentration in the center, lower concentration in the surrounding areas, and dispersion from the center towards the periphery, accompanied by an increase of -13.8% and 15.8% from 2001 to 2005 and 2010 to 2015, respectively; (2) The urban form factors of Beijing exhibited clear spatial and temporal evolution characteristics from 2000 to 2020, with average Total class area, Largest patch index, and Effective mesh size experiencing growth rates of 56.4%, 78%, and 112%, respectively. There was also a decreasing trend in average Patch density, while changes in Patch cohesion index and Interspersion & Juxtaposition index were found to be insignificant; (3) A significant correlation existed between urban form and carbon emissions, with Patch Density and Effective mesh size showing negative correlations, while Patch cohesion index exhibited a positive effect on carbon emissions; (4) The hierarchy of urban form factors influencing carbon emissions was as follows: Effective mesh size > Total class area > Largest patch index > Patch cohesion index > Interspersion & Juxtaposition index > Patch density > Perimeter-area ratio, moreover, the interaction between Effective mesh size and Interspersion & Juxtaposition index exhibited the greatest impact on the spatial heterogeneity of carbon emissions. Various regions should assess how urban form affects carbon emissions based on local conditions and subsequently develop strategies to optimize their spatial patterns.