Decisions on air quality problems must often be made on the basis of existing ambient air quality data. One consideration in such situations is how to accomodate the uncertainty associated with these measurements. Measurement error is often stated in terms of a single measurement, while the decision is made on the basis of the entire data set for a year. Therefore, there is a practical need to translate the uncertainty statement for a single measurement into an uncertainty statement for the decision-making statistic. In a sense, the quality assurance statement for an individual measurement should be transformed into a quality assurance statement that is applicable to the decision that is made from the data. This paper develops an empirical Bayes framework for this type of problem and examines the application of this technique for a short-term National Ambient Air Quality Standard. The analytical results obtained from the empirical Bayes methodology are compared to results obtained from computer simulations.