Small scale melt-mixing of polypropylene (PP) with 0.5 -7.5 wt% multiwalled carbon nanotubes (MWCNT) Nanocyl NC 7000 was performed. A comparison was made between the efficiency of masterbatch (MB) dilution and direct nanotube incorporation on nanotube dispersion and electrical percolation on three PP types differing in the molecular weight. In addition, two maleic anhydride (MA) modified PPs were used. To compare the two incorporation procedures, the specific mechanical energy (SME) during processing in a DSM Xplore 15ccm microcompounder was recorded and hold comparable by adapting the mixing time in direct incorporation performed at 210°C and 250 rpm. Masterbatches with 7.5 wt% MWCNT were mixed and diluted for 5 min each; the comparison of both methods was performed on compression molded plates at 2 wt% MWCNT. Based on optical transmission microscopy investigations, the MB dilution technique resulted in better MWCNT macrodispersion than the direct incorporation, independent of the PP viscosity and MA modification. Interestingly, despite better dispersion the electrical resistivity of samples with 2 wt% MWCNT was higher when using the MB approach, indicating a possibly more pronounced MWCNT shortening in this two-step procedure. In accordance to previous literature on PP, with increasing matrix viscosity electrical percolation threshold increased. However, in contrast to previous results on other polymer matrices the macrodispersion of the MWCNTs improved with decreasing matrix viscosity. This indicates the importance of melt infiltration into the primary agglomerates as first step of the dispersion process which is more pronounced at lower matrix viscosity. In summary, this study on PP implies that the MB approach is very suitable for the dispersion of MWCNTs and results in advantages for nanotube dispersion. However, electrical resistivity may be reduced at better dispersion and higher SME input.