Background and Aim: Mobile platforms can capture the hyperlocal variation of air pollutants in a complex urban terrain, in a limited time, and with a limited number of (costly) high-resolution monitoring devices, but limited temporal coverage per location. In this practical review, we focus on the applicability of mobile monitoring to develop air pollution maps for use in epidemiological studies. Methods: We assessed a wide range of mobile monitoring studies and summarize findings and conclusions, focusing on the critical design issues. We highlight the trade-off between the temporal and spatial variability, as well as differences between pollutants. We specifically assessed spatial coverage, number of streets and repeats per street, on-road versus of-road measurements, need for a reference site and statistical modeling options. Results: Robust mobile LUR models can be made with about 10% of the streets in the domain and limited (or no) repeats. Street segments with similar characteristics serve as pseudo repeats, meaning LUR models can be developed based on street segments with mobile measurements only measured once, if there is enough spatial and temporal coverage by including all spatial characteristics of the domain and measure during different parts of the day, days of the week and season. This also means that there is no critical need to temporally correct all measurements. To retain the local variation as much as possible, it is advised to keep the spatial resolution as low as possible (< 200m). It is possible to add data-only mapping on top of a LUR model in a mixed-model framework. Here, a LUR model is used to create a base map, and with more measurements more local variation can be added to the map. Conclusions: Mobile monitoring is a cost-effective scalable approach to map air pollution at fine spatial resolution.