As various countries seek to improve air quality by enforcing vehicle fleet regulation measures, better understanding the determinants of household car ownership and car type choice is critical. This would provide better inputs for agent-based mobility models, which can compute traffic-related daily emission profiles based on a synthetic population and a synthetic vehicle fleet. Usually, car type choice and ownership are estimated from household characteristics using discrete choice models (Jong and al. 2004). But recent studies point out the contribution of machine learning methods for the estimation of car ownership (Paredes and al. 2017; Dixon and al. 2021). This work investigates the performance of several classification models in the prediction of car type and ownership at the household level. We compare a discrete choice model against various machine learning classification methods (e.g. Gradient Boosting, Random Forest) for the estimation of household car ownership, fuel type and car pollutants emissions standards. Explanatory variables include household socio-economic characteristics as well as local and metropolitan accessibility variables (parking availability, public transport accessibility). The methodology is applied to the Paris area, using the “Enquête Globale de Transport” mobility survey. Considering the Matthew Correlation Coefficient, F1 score and Cohen’s kappa as evaluation metrics, we conclude that logistic regression slightly outperforms AI models for car ownership whereas Gradient Boosting classifier gets the best results for vehicle type estimation. Our results show a strong relationship for car ownership prediction and a slight agreement for fuel type and emission standard predictions, with a major importance of household composition and accessibility variables. This work emphasizes that discrete choice and AI models should rather be brought together than opposed to improve the performance of vehicle fleet synthesis in agent-based modeling.