Based on cross-sectional data from 30 Chinese provinces from 2004 to 2017, this paper systematically examines the nonlinear effects of economic policy uncertainty (EPU) on carbon emissions and its causes using the PSTR model. It is found that the impact of EPU on carbon emissions at the provincial level in China has significant nonlinear characteristics and shows a positive and then negative pattern as the level of EPU increases. Furthermore, increased levels of EPU also cause a nonlinear migration of the effects of provincial economic and financial development, industrial structure, government spending, and environmental regulation on carbon emissions, illustrating a large amount of heterogeneity among Chinese provinces. Specifically, provinces with higher levels of economic and financial development experience a greater positive carbon emission effect from EPU, whereas provinces with lower levels of such development experience a greater negative carbon emission effect. In contrast, in provinces with irrational industrial structures, lower fiscal expenditures, and weaker environmental controls, the nonlinear carbon emission consequences of EPU are greater. Therefore, local governments should prudently adjust economic policies, improve and perfect the market information disclosure system, and afford full play to regional comparative advantages to help achieve the “double carbon goal”.