Recent experiments indicated that the step mountain representation used in the NCEP Meso Eta Model may have difficulty in reproducing downslope windstorm at higher resolutions. However, earlier studies also showed that the use of the terrain following sigma coordinate may have contributed to forecast errors in the event of lee cyclongenesis. Tests have begun which will examine the performance of the sigma vs. eta modes in the NCEP Meso Eta Model. The first case study chosen is one in which both sigma-based NGM and AVN Models failed to capture the effect of lee cyclongenesis in their 48 h forecast and predicted cyclones that were too far north of its actual location. The operational Meso Eta Model, on the other hand, provided a much better forecast. The simulation results of the sigma mode in the Meso Eta Model show a lot of similarities to the operational Meso Eta run in both mass and momentum fields. The surface low in the sigma run of the Eta model, unlike the forecast in the NGM and AVN, is almost as close to the actual location as the one in the operational Meso Eta run. This suggests that the major reason for the differences between the operational Meso Eta Model forecast and the NGM and AVN forecasts was not the representation of mountains. More case studies regarding the comparisons between using these two vertical coordinates in the Meso Eta Model will be presented.