1. Urban areas affect terrestrial ecological processes and local weather, but we know little about their effect on aerial ecological processes. 2. Here, we identify urban from non-urban areas based on the intensity of artificial light at night (ALAN) in the landscape, and, along with weather covariates, evaluate the effect of urbanization on flight altitudes of nocturnally migrating birds. 3. Birds are attracted to ALAN, hence we predicted that altitudes would be lower over urban than over non-urban areas. However, other factors associated with urbanization may also affect flight altitudes. For example, surface temperature and terrain roughness are higher in urban areas, increasing air turbulence, height of the boundary layer, and affecting local winds. 4. We used data from nine weather surveillance radars in the eastern US to estimate altitudes at five quantiles of the vertical distribution of birds migrating at night over urban and non-urban areas during five consecutive spring and autumn migration seasons. We fit generalized linear mixed models by season for each of the five quantiles of bird flight altitude and their differences between urban and non-urban areas. 5. After controlling for other environmental variables and contrary to our prediction, we found that birds generally fly higher over urban areas compared to rural areas in spring, and marginally higher at the mid layers of the vertical distribution in autumn. We also identified a small interaction effect between urbanization and crosswind speed, and between urbanization and surface air temperature, on flight altitudes. We also found that the difference in flight altitudes of nocturnally migrating birds between urban and non-urban areas varied among radars and seasons, but were consistently higher over urban areas throughout the years sampled. 6. Our results suggest that the effects of urbanization on wildlife extend into the aerosphere, and are complex, stressing the need of understanding the influence of anthropogenic factors on airspace habitat.