Obtaining realistic meteorological fields that are necessary for numerical modelling and prediction of sea state within operational oceanography as well as for Rapid Environmental Assessment (REA) tasks is not easy in the regions like the Adriatic Sea, which is surrounded by complex local orography. By analysing the output fields from the mesoscale meteorological model LAMI and the global ECMWF model between October 2005 and September 2006, this paper investigates effects that orography may exert on meteorological fields over the Adriatic region. The analysis is performed over four weather regimes typical for the Adriatic Sea, namely those characterized by bora, sirocco, boraa??sirocco and etesian winds. Analysed are mean fields and mean differences between the two models. Results indicate that all investigated flow regimes are significantly affected by the Apennines and the Dinaric Alps, but also by smaller scale orography that is, at least along the Italian coast, adequately represented by LAMI and not by the ECMWF model. A kind of symmetry is found between sirocco and etesian regimes on the Adriatic scale, as well as on smaller scales, e.g. around the Gargano peninsula which acts as a typical 3D obstacle. Well known jets-and-wakes bora pattern along the eastern coast is clearly seen in the mean fields. The enhanced cloudiness and precipitation regions are shown to appear in convergence zones of bora, boraa??sirocco and sirocco regimes, related to the mountain ridges perpendicular to the respective flow. Additionally, within the boraa??sirocco regime a relatively narrow convergence zone at the boundary between the bora and sirocco flow is noticed. The authors believe that methods used in this paper can be useful when applied to other coastal regions and models. By considering the differences between output fields from two models of different resolution, the areas affected by smaller scale phenomena could be clearly seen. On the other side, the mean modelled fields show persistent meso to synoptic scale structures thus enabling better understanding of dynamical and climatological characteristics, both of the considered region and the models used.