We evaluate nitrogen dioxide (NO2) simulations from a widely used air quality model, the Environmental Protection Agency (EPA) Community Multiscale Air Quality (CMAQ) model, using ground‐ and satellite‐based observations. In addition to direct comparison of modeled and measured variables, we compare the response of NO2 to meteorological conditions and the ability of the model to capture these sensitivities over the continental U.S. during winter and summer periods of 2007. This is the first study to evaluate relationships between NO2 and meteorological variables using satellite data, the first to apply these relationships for model validation, and the first to characterize variability in sensitivities over a wide geographic and temporal scope. We find boundary layer height, wind speed, temperature, and relative humidity to be the most important variables in determining near‐surface NO2 variability. Consistent with earlier studies on NO2‐meteorology relationships, we find that, in general, NO2 responds negatively to planetary boundary height, negatively to wind speed, and negatively to insolation. Unlike previous studies, we find a slight positive association between precipitation and NO2, and we find a consistently positive average association between temperature and NO2. CMAQ agreed with relationships observed in ground‐based data from the EPA Air Quality System and the Ozone Monitoring Instrument over most regions. However, we find that the southwest U.S. is a problem area for CMAQ, where modeled NO2 responses to insolation, boundary layer height, and other variables are at odds with the observations.