Deteriorated air quality in cities worldwide is a current problem city administrations face. With the growing mobility requirements of the population living in cities, the number of registered cars and traffic intensity keeps increasing. Numerous studies emphasize the significant contribution of anthropogenic activity to deteriorated air quality. Initiatives linked to promoting bicycle transport, electromobility, or reducing car traffic as the primary source of pollution in urban areas are beginning to appear. However, simple solutions such as adding bicycle lanes or widening existing roads are often not applicable in European cities with heritage-protected historic centers. Modern numerical models are becoming mature enough to serve as a basis for reliable tools providing quantification of the impact of automobile traffic and significant anthropogenic sources in high-fidelity resolution - up to metre scale. However, several important limitations resulting from high-resolution processes need to be considered. First of all, models require very detailed input data about air quality. This includes the transport of pollution from outside the area of interest, emissions from local sources, or resuspension. Second, simulations using numerical models need to be properly validated. Particulate matter, as well as nitrogen (di-)oxide, are highly variable both in space and time. Finally, air quality strongly depends on meteorological conditions. Air quality models typically use national forecast models or regional climate models as meteorological drivers. Our results suggested that currently, no one model can be considered “best” for all potential applications. It seems to be beneficial to test at least two or three different models to analyse potential over- or under-estimation of the results. Practically, the article aims to point out the complexity of measuring and modelling air pollution at micro-scale resolution and uncertainties in fine-scale model resolution.