Addressing the challenges posed by pollutants is necessary to achieve the Sustainable Development Goals (SDGs) target 12 or achieve sustainable production and consumption patterns. Convergence assessment of air pollution provides information which can be beneficial to how to handle that air pollution across different countries. Nitrogen oxides (NOx) are one of the most popular air pollutants. However, the current empirical literature on environmental economics largely ignores the convergence of per capita NOx. For this reason, this study investigates the stochastic convergence of aggregate and fuel-specific per capita NOx emissions in 20 OECD countries. This paper employs a recently introduced panel stationarity test that considers both smooth and sharp structural changes in the data generation process. The panel results show that the convergence hypothesis is rejected only for NOx from light-oil consumption. However, country-specific results reveal substantial evidence for divergence in the sample countries when NOx emissions per capita generated through diesel consumption, light-oil consumption, and natural gas consumption are considered. Besides, we find that most of the series have convergent behaviour for aggregate NOx, NOx from biomass consumption, NOx from hard-coal consumption, NOx from heavy-oil consumption, and NOx from the process. The policy implications of the empirical results for proper environmental management are elucidated in the paper. Actions taken based on the convergence findings will likely lead to a decrease in NOx emissions per capita as the countries will converge towards a lower level of NOx emissions per capita, in line with SDG's target 12.