Abstract A generalized urban air pollution model, based on numerical integration of the concentration equation, is developed for the study of air pollutant distributions over an urban area. The model computes the temporal and three-dimensional spatial concentration distributions resulting from specified multiple point and area sources by using currently available meteorological and source inventory data. A new method based on experiments and a turbulence transport model is used to estimate the turbulent diffusivity and atmospheric stability. Special treatments of the finite-difference scheme to accommodate the large variations of concentrations are discussed. An effort has been made to avoid any subjective analysis scheme for the preparation of the input data. The model was used to study SO2 distributions in the St. Louis metropolitan area during 25 consecutive days in February 1965. The computed results were evaluated with respect to observed data by using various statistic methods. The computed results ...