Abstract In this study we developed an exploratory method to investigate the potential effect of VOC reduction from specific source categories on maximum ozone. The method is based on measurement of speciated organic compounds in air and application of chemical mass balance receptor modeling (CMB) to allocate ambient concentrations to specific source categories. This evaluation results in a sample-specific emission inventory, an estimate of the VOC emissions by source in an air mass which includes variations in source discharge not reflected in the conventional emission inventory. In addition, the CMB model in conjunction with the chemical profiles for each source, provides a basis for allocating very reactive organic components not measured in the atmosphere but contributing to O 3 formation. The carbon bond reaction kinetics model (CBM-IV) is used with observed meterological conditions and the specific source contributions to predict precursor conversion to ozone. Predictions of maximum O 3 using the technique for 9 days in Chicago in the summer of 1987 compared favorably with measured ground-level concentrations determined to be downwind by trajectory analysis. Source categories modeled included vehicle tailpipe exhaust, gasoline vapor, petroleum refinery emissions, architectural coatings, graphicarts, vapor degreasing, dry cleaning, and waste water treatment. The end-product is a family of control response curves which show the relationship between changes in VOC precursor emissions from each source and potential maximum ozone. The control functions for vehicle exhaust indicated that control of VOC from automobiles will be effective on some high O 3 days but not on others. The method complements grid-based photochemical models in that it will allow many more control options to be investigated than would be possible due to computational constraints.