The United States Environmental Protection Agency (U.S. EPA) is extending its Models-3/Community Air Quality (CMAQ) Modeling System to provide detailed gridded air quality concentration fields and sub-grid variability characterization at neighborhood scales and in urban areas. CMAQ is an advanced air quality modeling system that embodies a 'one-atmosphere,' multiple-pollutant philosophy. The meteorological model used with CMAQ in this application is the Pennsylvania State University/National Center for Atmospheric Research (PSU/NCAR) Mesoscale Model (MM5). For fine-scale urban simulations (approximately 1-km grid spacing), MM5 is being modified to include an urban canopy parameterization that accounts for the drag exerted by the urban structures on the flow, the enhancement of turbulent kinetic energy especially near the top of the buildings, and the energy budget at the street and roof levels. This refinement is targeted to provide CMAQ with the means to capture the details of the pollutant spatial distributions at these scales. One of the goals of this research is to demonstrate the capability of MM5 to simulate the effects of urban areas at the meso-gamma scale. Preliminary experiments with the 'off-the-shelf' MM5 were performed with various physics options in the coarser domains (e.g., 36, 12, and 4 km) to provide a baseline for comparing with the 1.33-km simulations. This paper will show sensitivities of the predicted meteorological variables to grid sizes and illustrate the utility of the urban canopy parameterization through comparisons with measurements. Ultimately, the urban canopy parameterization will be expanded to use more detailed land use databases than currently available in MM5, and it will include better-resolved urban morphology to identify urban terrain zones.