Introduction: The study aims to select the optimal model to estimate the emissions from vehicles with compressed natural gas (CNG) powertrain systems in Tehran to identify, evaluate and plan for controlling emissions through CNG vehicles, and determining the pollution level of passenger cars produced by the Iranian automotive industries and compare it with pollution standards. Material and methods: In this study, locally-produced CNG vehicles' information was used to estimate the pollutants emitted from domestic CNG vehicles using IVE and COPERT modeling software. Finally, the modeling outputs were compared with the results of the ISQI emission test. Result and discussion: Results showed that contrary to the initial assumption, supposing IVE software was suitable for advancing the research objectives, the performance of IVE software was not reliable for CNG vehicles, and the outputs of COPERT software were much closer to the values measured in the ISQI emission test. The COPERT estimated values were more accurate by 65 and 78 percent for carbon monoxide and NOX, respectively, but 13% worse for carbon dioxide. Conclusion: Based on the findings, the COPERT model is more accurate than IVE for modeling CNG vehicles' emissions, especially for CO and NOX. According to the findings, it is necessary to reconsider the superiority and application of IVE software for determining the emission inventory and estimating the air pollutants emitted from the CNG fleet in the country.