The Community Multiscale Air Quality (CMAQ) model is a state-of-the-science chemical transport model (CTM) capable of simulating the emission, transport and fate of numerous air pollutants. Similarly, the Weather Research and Forecasting (WRF) model is a state-of-the-science meteorological model capable of simulating meteorology at many scales (e.g. global to urban). The coupled WRF-CMAQ system integrates these two models in a “two-way” configuration which allows feedback effects between the chemical (e.g. aerosols) and physical (e.g. solar radiation) states of the atmosphere and more frequent communication between the CTM and meteorological model than is typically done in uncoupled WRF-CMAQ simulations. In this study we apply the various cumulus parameterization (CP) options available in WRF at horizontal grid spacings ranging from regional scale (i.e. 12-km) to urban scale (i.e. 4 and 1 km), focused on the July 2011 DISCOVER-AQ campaign that took place over the Baltimore-Washington D.C region. Of particular interest is the evaluation of the WRF simulated clouds, as analysis of previous WRF-CMAQ simulations using a “standard” 12-km configuration for the model suggest that WRF has difficulty predicting clouds (particularly fair-weather clouds), with decreasing skill at finer horizontal grid spacings. Here we will examine the impact that the WRF CP options have on cloud predictions, using available satellite data to evaluate model the performance. We then examine how changes in the WRF simulated clouds affect CMAQ predictions of ozone and PM2.5 at the various scales.