Regional air quality simulation models have been developed as research tools to gain a greater understanding of long range transport and transformation of air pollutants and to evaluate regional control strategy alternatives. To ensure credibility of these complex model predictions, the performance of model codes must be verified for their adherence to theoretical formulations, accuracy of numerical algorithms, ability to reproduce observed data and sensitivity to uncertainties in input data. The paper describes the procedures used for verification of the U.S. EPA Regional Oxidant Model. Comparison with observed data are presented together with a brief summary of the results from two ozone control strategy applications. The sensitivity of the model to known uncertainties in input data is discussed in relation to its effect on air pollution control strategy evaluation.