Different Data Assimilation techniques have been formalized and applied in the context of complex nonlinear models, to describe chemistry and physics of the atmosphere. In the literature the main approaches presented are based on a) statistical interpolation (SI) techniques, including optimal interpolation methods, residual kriging methods, regression, etc... and on b) variational methods, as well as ensemble methods such as Ensemble Kalman filters (EnKF). The aim of all these methods is to combine various data sources, to provide an optimal estimate of the spatial distribution of a particular pollutant, considering the uncertainties in the measurements as well as in the models. This paper presents the Ensemble Kalman Filter (EnKF) scheme used to assimilate ozone measurements from ground monitoring stations in the simulations performed by an air quality model system. The Data Assimilation scheme has been applied to Northern Italy. Results show that the methodology highly improves the ozone estimation.