A linear programming model is used to estimate the cost of achieving certain air quality goals for the St. Louis airshed. The choice set of control methods is based on engineering data compiled by the author. The model is used to generate a set of alternative air quality levels for sulfur dioxide and particulates which can be attained at the same total cost as the goals initially considered. This frontier of trade-off possibilities is compared to a particular social indifference curve which is based on medical considerations. It is found that both curves are concave in the same direction.