As of July 2017, the emissions type-approval of light-duty vehicles in Europe is based on the Worldwide Harmonized Light-duty vehicles Test Procedure (WLTP), introduced to replace the old and outdated New European Driving Cycle (NEDC) test procedure. Since some elements of the European Legislation are still based on the NEDC (2020 CO2 emission targets, vehicle labelling, national vehicle taxation policies, etc.) in order to allow sufficient lead time to vehicle manufacturers and national authorities to adapt to the new procedure, a simulation-based approach was chosen to calculate CO2 emissions and fuel consumption according to the NEDC regime in the period 2017–2020. To achieve this objective without significantly increasing the cost and duration of the certification procedure, existing regulation foresees that vehicles are tested over the WLTP for CO2 emissions, the test results are used as input in a simulation model that then calculates the corresponding CO2 according to the NEDC test protocol. A dedicated vehicle simulation model (CO2MPAS) was developed for the purpose and is currently used for the type-approval of new vehicles in Europe. The development specifications of CO2MPAS were challenging, as it had to be highly accurate, exhibit fast operation, and function with a limited number of input data. This paper presents the development principles and process followed, details of the physical models employed in CO2MPAS, and provides information regarding its accuracy, validity and in use operation. CO2MPAS achieves low errors in the prediction of the NEDC cycle that in the controlled sample used for its development are of the order of 1% with a standard deviation of 3%, while the respective in-use numbers are of the order of 1.5% and 5%. In parallel, random sampling and testing of a 10% of the type-approved vehicles also occurs in order to guarantee the quality of the CO2MPAS results and the validity of the process. It is concluded that CO2MPAS can be used to accurately estimate emissions of conventional vehicles within a ±4% accuracy range, even when limited input data are available. In addition, the in-use data analyzed suggest that the use of the tool enables the certification of about 2/3 of the new vehicle models without the need of additional experimental tests. This is an important achievement as it reduces the costs and time necessary to certify light-duty vehicle CO2 emissions during the transitional period. Finally, it can be concluded that the use of CO2MPAS does not affect the declared CO2 emissions of vehicles over NEDC conditions.