This paper analyzes data obtained on July 3 and 18, 1995, during the summer 1995 Southern Oxidant Study (SOS) field campaign. In a previous paper <Nunnermacker et al., 1998> we analyzed measurements of key species that contribute to formation of O3 in the Nashville urban plume and presented a semiquantitative picture of O3 production in the plume from the point of emission to locations where no net O3 was being formed. In this paper we use a box model constrained by observed concentrations of stable species to obtain a detailed mechanistic description of the instantaneous processing of urban emissions at various times in the chemical evolution of the urban plume. Instantaneous ozone production rates and efficiencies with respect to NOx and to primary radical production are examined. At high NOx concentrations in the fresh urban plume the O3 production rate was found to be directly proportional to the hydrocarbon to NOx reactivity ratio. At lower NOx concentrations, corresponding to the mature urban plume and the background atmosphere, the O3 production rate was found to be directly proportional to the NOx concentration and independent of the hydrocarbon reactivity. NOx was found to be most efficiently used for ozone production at low NOx concentrations. In contrast, the efficiency with which the system uses primary radicals was found to be very low at low NOx concentrations and to peak at a NOx concentration of approximately 4 ppbv. A sensitivity study of the instantaneous O3 production rates to changes in NOx or hydrocarbon concentrations showed that the instantaneous O3 production rate at the center of the urban plume, when half of the urban NOx emissions had been processed, is hydrocarbon sensitive. However, O3 production becomes NOx sensitive as the plume matures. ¿ 2000 American Geophysical Union