The monitoring programs needed to design control strategies differ from those required to assess the impact of the strategies on air quality. The former are short-term and intensive. They are designed to obtain the data required to apply an urban-scale grid model, like the Urban Airshed Model, or a combination of an urban-scale and regional-scale model, like the Regional Oxidant Model, to an area without having to assume important input data. Such measurements include three-dimensional pollutant and meteorological measurements throughout the modeling domain. Detailed and accurate emissions inventories are also required. Model validation should compare not only the 03 predictions with actual data, but also include comparisons for NOx, individual volatile organic compounds (VOC), and if possible, additional species such as PAN, H202, formaldehyde, and HNO 3. To assess the impact of control strategies, long-term monitoring networks that measure 03, NO x and speciated VOC are needed. 03 trends should be examined using robust, meteorologically-adjusted 03 statistics. Such statistics need to be related to a robust National Ambient Air Quality Standard.