This paper applied the nonlinear autoregressive distributed lag (NARDL) model to investigate the nonlinear asymmetric cointegration relationships between trade liberalization and greenhouse gas emission (GHG) in Iran. Also, the methane gas emission per capita was used as greenhouse gas emission index. The annual data from 1982 to 2013 was used. The bounds test of the NARDL specification suggests the presence of co-integration among the variables, which includes the methane emission per capita, trade liberalization, energy consumption per capita and GDP per capita. Furthermore, the analysis indicated that there is a long-run asymmetric relationship between trade liberalization and methane emission in Iran. But, the short-run relationship is symmetric. The long-run estimation showed that the positive and negative shocks arising from trade liberalization increase methane emissions by 0. 22 and 0. 9 percent, respectively. The empirical results revealed that the relationship between methane emission and GDP per capita in long and short run is N-shaped and statistically significant. Also, energy consumption has a significant positive effect on methane emissions and 1 percent increase in per capita energy consumption leads to 0. 75 percent growth in methane gas emission.