The identification of static and dynamic behavior (properties) is an important step in design and development of cars. However, the time necessary to calculate and evaluate dynamic performance appears to be increasing. This is firstly due to a rise in the number of degrees of freedom in the model, caused by ever-increasing accuracy in the physical reproduction of car details. It is also due to the large number of possible design variants (concepts), which have to be calculated and compared before the optimum solution can be selected. Due to ever shuttering product design cycles, the development process is increasingly carried out on a virtual basis. Therefore it is often necessary to make design decisions based on the results of FE models, which must be validated by comparison with other FE models or experimental results. The article describes some new procedures for the identification of modes, the comparison of calculated and measured dynamic results, and model updating of large FE structures. Numerous car bodies as well as power train structures are subjected to a comprehensive set of experimental and analytical mechanical engineering methods. The correlation between the results from the different methods is discussed extensively. Different FEM models results are checked and the analytical structural modes are calculated using the software packages FEGraph and FEOptim, and MSC.nastran 2004 and PEMAS packages for the numerical solutions. These FEM results are then utilized in pre-test analysis tools to validate the excitation and response locations for the experimental modal analysis. A full sequence of experimental modal measurement results is obtained and the results are correlated with analytical FEM model results. The examples show that the modal testing technology should be a basic tool for all NVH FE engineers. The use of the available tools and the user programming capabilities of FE packages like MSC.nastran, PERMAS, etc. provide an unique environment that guides the engineer through the complete process, from pre-test analysis over correlation to end up eventually at the model updating step. The interpretation of the calculations can be visualized. All modules of post-processing also integrate test modules, which could support requirements of industrial large FE models.