Diagnostic evaluation of the Eta/CMAQ air quality forecast system has shown generally excellent performance for ozone prediction under clear sky conditions and much poorer performance under cloudy or rainy conditions. Degraded ozone predictions are to be expected under cloudy conditions, and are probably inevitable to some degree, because of difficulties in accurate prediction of cloud parameters relevant to photolysis and vertical mixing. We are working to improve parameterizations in CMAQ that use meteorological information to compute actinic flux for photolysis and convective transport of chemical species. This study involves diagnostic evaluation of model performance under conditions ranging from scattered fair weather cumulus to thick overcast and deep convection. Model modifications involving cloud effects on radiation attenuation and involving deep convective transport are described and tested. Preliminary sensitivity testing indicates that overprediction of ozone concentrations during cloudy conditions can be significantly reduced by modification of the cloud cover diagnostic algorithm used in the CMAQ system. In addition, we are testing a new technique using cloud attenuation factors derived from the Eta predictions of ground-level short-wave insolation. This technique has the advantage of greater consistency with radiation modeling in the meteorology model and the potential for improvements with advances in the Eta and WRF systems. The Eta/CMAQ system often shows large overpredictions of ozone associated with deep convective systems such as along cold fronts. Such overpredictions became more pronounced after the introduction of dynamic ozone lateral boundary conditions (LBC) derived from the Global Forecast System (GFS). The switch to GFS LBCs resulted in much higher ozone concentrations in the highest model layers (100 – 300 mb). CMAQ’s subgrid convective scheme simulates downward transport by downdrafts starting above the top of the convective cloud. Sensitivity experiments show that ozone overpredictions are greatly reduced when the simulated downdraft is turned off. Disclaimer - Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy