In this work, a complex numerical approach was developed to predict electrical conductivity response of polymer composite materials to uniaxial deformation. Composites were based on thermoplastic polymer – polypropylene (PP) reinforced with nanoscale carbon filler of two types: multi-walled carbon nanotubes (MWCNTs) and carbon black (CB). Numerical predictions were compared against experimentally obtained results. Numerical modeling was conducted using Finite Element Method (FEM) and embedded element technique, allowing computationally efficient way to simulate behavior of systems with high number of particles (up to 15 vol.% of CB particles and 7.5 vol.% of MWCNTs under relatively large deformations (up to 10% of strain). Effect of the particles agglomeration on the response of the material’s electrical conductivity due to applied load was investigated for different agglomeration degrees of both types of the filler using the concept of Representative Volume Element (RVE). To verify developed numerical method, a series of specimens of PP-based composites were manufactured using melt mixing method. To obtain different degrees of agglomeration, a fine PP reactor powder was used. PP powder mixed with nanoscale filler was subjected to ultrasonication in a volatile liquid with different treatment times, followed by melt-mixing of the dried compounds. Electron microscopy was used to control agglomeration degree of the samples. Resulting composite materials in a form of thin plates were subjected to uniaxial deformation, while electrical conductivity response was constantly measured. The obtained dependencies were compared to the ones obtained numerically. It was concluded, that the developed numerical approach provides satisfactory level of accuracy predicting electrical conductivity changes of the polymer composites under deformation. The reported study was funded by RFBR according to the research project № 18-33-00688.