precursor levels. Extending the traditional approach of using domain-wide maximum ozone values as the metric for guiding the development of emission control strategies, the new metrics incorporate attributes of the spatial and temporal pervasiveness and the severity of the ozone episodes considered for strategy development, as well as the impact on potential exposures to ozone. The usefulness of using various alternative criteria to better understand the directionality of the impact of emission controls is demonstrated via a set of 26 simulations of a three-day period of the severe July 1988 episode over the New Jersey-Philadelphia-Delaware Valley area. These simulations model the effect of across-the-board reductions of VOC and NO by 25, 50, 75, and 100% of the base case. The impact of these reductions is found to be dependent on the control objective (e.g., reduction of exposure vs. reduction of the maximum) as well as on the targeted level of the control objective (e.g., reduction of exposures below 120 ppb vs. reduction of exposures below 80 ppb).