Due to the rapid growth of fossil energy consumption, countries worldwide have paid considerable attention to reducing carbon emissions. Moreover, with economic globalization and trade liberalization, exploring the relationship between foreign trade and carbon emission reduction has become increasingly critical. Exploring this relationship can aid in establishing suitable recommendations for global carbon emission reductions. This paper uses a spatial econometric model and a dynamic panel threshold model to empirically test the spatial effect, nonlinear effect, and heterogeneous effect of foreign trade on global carbon emissions. All the above models are based on the construction of the economic weight matrix of different countries. The results reveal that 1) carbon emissions in various countries exhibit with significant spatial spillover in the overall spatial context; 2) foreign trade has a significant role in promoting carbon emissions in local and similar economic areas, but it has an apparent dual-threshold effect on economic development; and 3) there are significant differences in the impact of foreign trade on carbon emissions in different regions and different periods. Therefore, in the process of global economic integration, based on their development stages and comparative advantages, countries can focus on overall planning and coordination to promote the optimal allocation of resources and reduce carbon emissions.