Population concentration generates both economic benefits and environmental costs through agglomeration-related spatial externalities. This paper examines how population density shapes fine particulate matter (PM2.5) concentrations across U.S. states over the period 2000–2023, conceptualizing density as a structural driver of nonlinear environmental externalities within regional systems. Using a balanced panel of all fifty states, I estimate fixed-effects models in which logged PM2.5 is modeled as a flexible quadratic function of logged population density, allowing coefficients to evolve over time and across density regimes. The results reveal a convex relationship between population density and PM2.5, indicating that the environmental externalities of agglomeration increase at an accelerating rate over much of the density distribution. However, the strength of this nonlinear effect has weakened over time, consistent with regulatory adaptation and technological diffusion. A split-regime analysis further shows that the curvature of the density–pollution relationship differs systematically between low- and high-density states, suggesting heterogeneity in regional externality structures. These findings contribute to the regional science literature by demonstrating that the environmental consequences of population concentration are nonlinear, heterogeneous, and dynamic within an advanced economy context. More broadly, the results highlight the importance of incorporating spatial externalities and governance scale into density-sensitive regional policy design, particularly in transportation investment, land-use planning, and environmental regulation.