CO₂ is one of the crucial concerns of the global economy due to its negative environmental impact. Based on the limitations of extant studies, this research employs the STIRPAT framework to investigate the asymmetric effect of urbanization (URB), gross domestic product (GDP), and energy intensity (EINT) on carbon dioxide emissions (CO₂) in Brazil from 1980–2022 using innovative econometric techniques such as Non – linear Autoregressive Distributed Lag (NARDL) and Dynamic Ordinary Least Square (DOLS). The NARDL findings are as follows: (i) The NARDL bounds test confirmed the existence of long – term association; (ii) In the long – term, a positive change in URB reduces CO₂ by 8.6%, while an adverse change in URB does not affect CO₂. In the case of GDP, a positive change raises CO₂ by 1.73%, whereas a negative change reduces CO₂ by 2.94%. For EINT, a positive change induces CO₂ by 2.12%, while a negative change decreases CO₂ by 3.82%; (iii) In the short – term, an adverse change in URB spurs CO₂ by 0.39%. For GDP, a positive change raises CO₂ by 1.73%, whereas a negative change reduces CO₂ by 2.94%. Lastly, a positive EINT change drives CO₂ by 2.12%; (iv) The DOLS findings confirm the NARDL results. The policy implications are discussed.