The Ozone DEposition Model (ODEM) was developed for national scale modeling of total and fractional ozone fluxes to grasslands and wheat using high-resolution gridded input data. The aim of this study was to improve the preliminary version 1.1 to enable season-long simulations by introducing a new parameterization of leaf area index (LAI), and the effect of soil moisture on soil surface resistance, and by refining the descriptions of friction velocity, u ; stability correction function, and all resistances depending on these parameters. Version 1.2 was tested by comparing simulation results with field measurements over grassland at Braunschweig, Germany, and wheat at Comun Nuovo, Lombardy, Italy. These comparisons showed that the simulation of ozone flux and u for grassland was greatly improved by the modifications, yielding a root mean square error for model vs. simulation of 0.056 ppbms 1, as compared to 0.106 ppbms 1 obtained with the original model. Atmospheric, boundary layer, and in-canopy resistances, which all depend on u ; increased as a result of the combined modifications. For wheat, the agreement between simulated and measured ozone flux was good under wet conditions, but less during a phase characterized by drying soils. It is suggested that in-canopy processes, which are uncertain in the model, become more important under these conditions. The application of the model using 6-year ozone and meteorological data from a site in southern Switzerland (Cadenazzo) revealed that the monthly cumulative stomatal ozone flux for grassland remained more stable across the growing season than for wheat because of less variation in green LAI. Highest stomatal ozone uptake in wheat occurred during anthesis and declined later due to decreasing green LAI. In both grassland and wheat, cumulative seasonal ozone uptake was less variable between years than was the AOT40 index. The results demonstrate the improvements in ODEM 1.2, and they underline the importance of species-specific parameterizations of plant development. But it is also concluded that for high-resolution modeling of ozone fluxes the overall quality of model outputs may critically depend on the quality of an extensive set of site-specific biological, climatic, and edaphic input data.