COVID-19 has raised new challenges for transportation in the post-pandemic era. The social distancing requirement, with the aim of reducing contact risk in public transit, could exacerbate traffic congestion and emissions. We proposed a simulation tool to evaluate the trade-offs between traffic congestion, emissions, and policies impacting travel behavior to mitigate the spread of COVID-19 including social distancing and working from home. Open-source agent-based simulation models were used to evaluate the transportation system usage for the case study of New York City. A Post Processing Software for Air Quality (PPS-AQ) estimation was used to evaluate the air quality impacts. Finally, system-wide contact exposure on the subway was estimated from the traffic simulation output. The social distancing requirement in public transit was found to be effective in reducing contact exposure, but it had negative congestion and emission impacts on Manhattan and neighborhoods at transit and commercial hubs. While telework can reduce congestion and emissions citywide, in Manhattan the negative impacts were higher due to behavioral inertia and social distancing. The findings suggested that contact exposure to COVID-19 on subways is relatively low, especially if social distancing practices are followed. The proposed integrated traffic simulation models and air quality estimation model can help policymakers evaluate the impact of policies on traffic congestion and emissions as well as identifying hot spots, both temporally and spatially.