The impacts of autonomous vehicles (AV) on safety, energy and atmospheric emissions have been recognised to be important issues, but an air quality impact assessment is missing. In this study, by using a numerical modelling approach, the impact of AV on the air quality of a medium-sized Portuguese urban area was evaluated. For that, the air pollutants nitrogen oxides (NOx) and carbon dioxide (CO2) were considered and three scenarios were developed: i) a baseline scenario; ii) an autonomous scenario, assuming an AV market penetration rate of 30%; and iii) an electric autonomous scenario, taking into account that those 30% of AV are pure battery electric cars. A modelling system composed by a road traffic model, a road transport emission model and a Computational Fluid Dynamics air quality model was used. The autonomous scenario promoted an increase of both NOx (+1.8%) and CO2 (+0.7%) emissions, while the electric autonomous scenario resulted in emission reductions of about 30% for both air pollutants. In terms of air quality, distinct patterns were found: i) the autonomous scenario promoted both increases and decreases of NOx concentrations; and ii) the electric autonomous scenario promoted a widespread reduction of NOx concentrations (with an average value of −4%). Overall the results showed that AV have the potential to improve urban air quality, but, further research is needed to enrich the findings of this work. Display omitted •Autonomous vehicles technology is seen as an opportunity to deal with cities air pollution.•Two autonomous vehicles scenarios were studied using the VISSIM-VSP-VADIS modelling setup.•The autonomous scenario promoted both increases and decreases of NOx concentrations.•The hybrid scenario promoted a widespread reduction of NOx concentrations.•Positive and negative effects were observed, showing the need of knowledge-based planning of traffic management.