This study attempts to investigate how economic growth (EG), energy consumption (EC), and population (POP) hurt the environmental quality of five regions: South Asia, East Asia, Latin America and the Caribbean, North America, as well as the Middle East and North Africa. The Wald and NARDL bounds tests check asymmetry and cointegration among the variables, respectively. The study has used the panel non-linear autoregressive distributed lag (PNARDL) model to analyze the non-linear panel cointegration and the panel short and longrun associations among the variables. In the long-run, EG with a negative shock has a positive and significant impact on CO2 emissions for East Asia and Latin America and the Caribbean. In the Middle East and North Africa, EG with a positive shock has a positive and significant impact on CO2 emissions. In North America, a positive shock in the EG has a negative and significant impact, while the negative shock positively impacts CO2 emissions.There is no significant impact of the decomposed EG in South Asia on the carbon emissions. Thus, the EC has a positive and significant impact on the CO2 emissions in all the regions except the Middle East and North Africa. The POP is also directly proportional to the CO2 emissions in all the regions. The results of the PNARDL show that in the longrun, the decomposed EG with positive shocks has a negative association, whereas the adverse shocks have a positive association with CO2 emissions.