This study investigates the relationship between geopolitical risk (GPR) and carbon emissions (CE) in the United States using monthly data from January 1985 to November 2025. Employing Vector Autoregression (VAR) and Nonlinear Autoregressive Distributed Lag (NARDL) models, we examine both linear and asymmetric effects of geopolitical risk on carbon emissions. Our empirical results reveal that geopolitical risk and carbon emissions exhibit a weak negative correlation over the full sample period, with the correlation coefficient being -0.11. However, this relationship varies significantly across different time periods, ranging from -0.29 during 1985-2000 to -0.17 during 2001-2008. The VAR model analysis indicates that while GPR shocks have limited immediate impact on carbon emissions changes, the dynamic responses show complex patterns over time. Granger causality tests suggest no significant causal relationship between GPR and carbon emissions in either direction. The NARDL model results demonstrate that both positive and negative GPR shocks have insignificant effects on carbon emissions, suggesting that the impact of geopolitical risk on carbon emissions is symmetric but statistically weak. These findings contribute to the ongoing debate on the environmental consequences of geopolitical tensions and provide important policy implications for climate governance in an era of heightened geopolitical uncertainty.