Microscale dispersion models resolve concentrations at spatial scales of meters over a domain of kilometers. Concentration estimates from such models have to be combined with ‘background' concentrations from regional scale models to obtain values that correspond to observations. In principle, these background concentrations can be obtained from Eulerian grid models, such as CMAQ. However, the computational demands of grid models can be prohibitive if concentrations are required for simulation times of the order of a year. This paper presents an alternate approach to modeling background concentrations of NOx, NO2, and O3. It is based on a receptor based Lagrangian model in which back trajectories are used to trace the history of an air parcel arriving at a receptor. The chemical species emitted into the air parcel are first transported without accounting for chemistry. However, each species is associated with an effective age in addition to concentrations. The species are then allowed to react in a box over a time corresponding to the maximum of the ages of the species that arrive at the receptor. This separation of transport and chemistry reduces the computational requirements of chemistry to allow simulation over time scales of years at multiple receptors. The paper presents the formulation of the model and the evaluation of concentration estimates from the model with observations made at 21 monitors covering greater Los Angeles.