Abstract A statistical model has been developed for the prediction of changes in oxidant air quality as a function of location resulting from the creation of an indirect source or the implementation of a land use or transportation plan. The model is implemented, as an illustrative example, for the regional air basin comprised of San Diego County, in the United States, and Tijuana, in Mexico. Los Angeles and San Diego geocoded reactive hydrocarbon emission inventories from mobile and stationary sources for 1969 and 1970 have been used with wind data and observed oxidant air quality in order to calibrate a statistical relationship between emissions and air quality. This relationship, applied to the San Diego-Tijuana region, has then been used with a 1975 base case geocoded emissions inventory to generate a base case 1975 oxidant index as a function of position. An altered 1975 geocoded emissions function incorporating the predicted emissions effects of changed land use can then be employed to generate a new 1975 oxidant function. The difference between the new and old oxidant functions represents the effect, area by area, of the change in land use on air quality. To demonstrate the possibilities of this type of model, three examples of air quality impact assessments have been carried out: first, the impact of a proposed regional shopping center, second, the effects of a proposed large scale development, third, an analysis of an “Existing Trends” 1995 alternative regional transportation and land use plan for the San Diego-Tijuana region. The resulting predicted changes in oxidant air quality, measured by the number of additional hours per year that the U.S. Federal oxidant standard is to be exceeded, are displayed as a function of position over the air basin. Given the emissions resulting from changed land use, the calculation of air quality change requires only a modest amount of computer time, and thus this type of model can inexpensively and rapidly be used to evaluate the approximate air pollution effects of planning alternatives.