More and more components made of short-fiber reinforced plastics are being used in modern powertrains. A reason for this are the good acoustic properties due to the lower stiffness and higher damping compared to classic metallic materials. In order to meet the increased customer demands regarding the acoustic sound comfort of internal combustion or electric powertrains, it is necessary to precisely predict the vibration behavior of components that are responsible for the transmission of structure-borne noise into the vehicle structure. However, at present the simulation cannot satisfactorily predict the actual vibration behavior of short-fiber reinforced plastics. The required material data, in particular damping, have so far often been obtained from static tests or are completely unknown. However, dynamic, frequency-dependent material properties are required for a reliable prediction of the special viscoelastic properties of the short-fiber reinforced plastics. By means of a new test method based on flexural resonance vibrations, viscoelastic material data can be characterized in a frequency range between 100 Hz and 10 kHz, taking into account environmental conditions such as temperature and humidity. Using these material data, a simulation of the structural dynamic behavior can be performed using either of two modeling approaches: microscopic or macroscopic. The basis is the second-order orientation tensor of the fibers from an injection molding simulation. The microscopic modeling approach uses a two-step homogenization of the properties of the matrix, fiber and matrix-fiber-interphase followed by a spatial discretization into material databases. The macroscopic modeling approach instead uses a one-step homogenization based on directly measured viscoelastic material data with different fiber orientation. The focus of the presentation is a detailed description and direct comparison of the two modeling approaches of stiffness and especially damping. Furthermore, the necessary assumptions and compromises as well as challenges will be discussed to categorize the usage of the models in proper cases. For this purpose, cuboid test specimens and a structural component made of fiber-reinforced plastic are investigated. A short digression explains the underlying dynamic material characterization.