To improve the predictability of injection-molding simulation tools the material behavior has to be determined under industrial, process-near conditions. Therefore, it is necessary to realize the implementation of the pressure-dependent thermal conductivity in commercially available software tools, such as Moldflow MPI 6.1. However, only few studies on the pressure dependence of polymers are available. The main issue of this study is to measure the thermal conductivity of a semi-crystalline polymer (polypropylene) at various temperatures, different pressure levels and for different filler degrees. A wide variety of models was used to describe the dependence of pressure and filler volume fraction on thermal conductivity of polymer multi-phase systems. Pressure dependence in the melt state can very well be described by a model developed from Rides et al. Another model suggested by Zhong and Yang allows to calculate the polymer thermal conductivity for the melt state, assuming the thermal conductivity at ambient pressure and the Tait coefficients from pvT-data are known. The thermal conductivity shows a linear increase with pressure. An increase from 0 to 800 bar results in a 15% rise of thermal conductivity for unfilled polypropylene in the melt state. In the solid state, however, the polymer thermal conductivity increases about 21%. The rise in thermal conductivity due to polymer crystallization averages to approximately 25%. An increasing filler volume fraction leads to a raise of polymer thermal conductivity. The influence of fillers can be very well described by the parallel-model (melt-state) and the model of Maxwell (solid state). However, the predictability of the parallel-model diminishes with higher filler volume contents. All other applied models calculate too high values for the thermal conductivity of the two-phase systems. The injection molding simulation indicates that an increase of thermal conductivity influences the results of the maximum injection pressure, the minimum cycle time and the frozen layer fraction. However, the influence to be expected on the calculation of shrinkage and/or warpage could not be verified. These results may be caused by the insufficient possibilities to consider pressure dependent thermal conductivity of polymer melts in commercial simulation software.