The continued need for more rational and practical assessments of hydrocarbon reactivity for use in formulating control strategies has led to suggestions that OH radical reactivity be used as a criterion for predicting oxidant formation. In order for such a reactivity scale to have maximum utility, correlations between the rates of reaction of hydroxyl radicals with organics and rates and levels of ozone formation must be established. To this end, 9-h irradiations of NOx-hydrocar-bon-air mixtures consisting of an alkane, an alkene, and an aromatic were carried out in a 6400-L all-glass smog chamber. The “standard” mixture consisted of n-pentane, m-xylene, and trans-2-butene, since the rate constants were known both for their reactions, and that of at least one isomer, with the OH radical. In one case the respective isomers were substituted sequentially at the same initial concentration and the effect of differences in OH rate constants on O3 production for constant carbon number was examined. In a second set of runs, the initial concentrations of organics were changed so as to maintain constant the OH reactivity (i.e., [HC] × kOH). To a good approximation the rate of formation and yields of both O3 and PAN in the earlier stages (≤6-9 h) of irradiation depended primarily on the OH radical reactivity and to a much lesser extent on the amount of carbon present. © 1979, American Chemical Society. All rights reserved.