Teleworking has been identified as a potential key lever for reducing air pollution. Yet, evaluating the atmospheric outcomes of teleworking enhancing policies remains difficult, especially when official databases on telework, household equipment and car emissions are incomplete or nonexistent. Here we propose several techniques to efficiently assess the impact of an increase in teleworking rates, and to explore the resulting bias, in a typical medium-sized European metropolitan area where few data are available: Besançon, France. Population and cartographical data are introduced in an individual-based daily mobility simulation model. We then calculate the resulting emissions for twenty atmospheric pollutants, using three different methodologies that aim to compensate, with different precision levels, for the lack of accurate information regarding vehicle fleets. Our results confirms the efficiency of telework for reducing emissions, with an average reduction of −0.42% in emission for an increase of 1% in teleworking rate. The precision level of data used strongly impacts the estimated quantity of air pollutant emissions (up to a factor ten). Failing to correctly account for inequalities in teleworking rate and equipment between socio-occupational categories introduces strong bias in the results which may degrade the correct evaluation of environmental benefits of teleworking enhancing planning policies.