This paper presents an overview of earlier and current methods of modelling atmospheric dispersion, and proposes and evaluates a screening model for operation over the World Wide Web. The On Line Atmospheric Dispersion Model (OLADMO) is a quasi boundary layer parameterised Gaussian plume model with additional algorithms to account for plume rise, building wake effects and deposition processes. The Monin-Obukhov length boundary layer parameter is utilised to define six stability classes in order to determine atmospheric turbulence and stability, whilst new equations, derived from an intercomparison study of old and next generation dispersion models, are used to calculate the horizontal and vertical dispersion coefficients r y and r z . Using data from two field experiments in Copenhagen, Denmark and Lillestrøm, Norway, the model results from OLADMO are found to compare favourably with the results from several old and next generation dispersion models. As a consequence of the unique nature of the meteorological and location factors of the Lillestrøm experiment, all models struggled to represent the concentrations observed during the field study adequately. However, OLADMO was the best performing model in this case, with a mean normalised crosswind integrated concentration 13% closer to the mean observed concentration than its nearest competitor. Because the evaluation of the model was conducted with a limited dataset, several limitations and improvements to both the model and experimental procedure are suggested.