Lockdowns implemented during the COVID-19 pandemic were utilized to evaluate the associations between “social distancing policies” (SDPs), traffic congestion, mobility, and NO2 air pollution. Spatiotemporal linear mixed models were used on city-day data from 22 US cities to estimate the associations between SDPs, traffic congestion and mobility. Autoregressive integrated moving average models with Fourier terms were then used on historical data to forecast expected 2020 NO2. Time series models were subsequently employed to measure how much reductions in local traffic congestion were associated with lower-than-forecasted 2020 NO2. Finally, the equity of NO2 pollution was assessed with community-level sociodemographics. When cities’ most stringent SDPs were implemented, they observed a 23.47 (95% CI: 18.82–28.12) percent reduction in average daily congestion and a 13.48 (95% CI: 10.36–16.59) percent decrease in average daily mobility compared to unrestricted days. For each standard deviation (8.38%) reduction in local daily congestion, average daily NO2 decreased by 1.37 (95% CI: 1.24–1.51) parts per billion relative to its forecasted value. Citizenship, education, and race were associated with elevated absolute NO2 pollution levels but were not detectibly associated with reductions in 2020 NO2 relative to its forecasted value. This illustrates the immediate behavioral and environmental impacts of local SDPs during the COVID-19 pandemic.