Multiwalled carbon nanotubes (MWNT) with dimensions of 8-15 nm and 0.5-2 μm long were used to make a series of CNT films via vacuum filtration, spin coating and inkjet printing. The method of deposition as well as the substrate used (filter paper of different sizes, photopaper, glass) were found to have an influence on the arrangement of the nanotubes and their resultant electrical response. Using a combination of impedance spectroscopy, equivalent circuit modeling, and microscopy techniques, it was possible to describe in detail how the electrical properties change as a function of how the MWNTs are distributed on the porous substrates by varying the number of deposited layers (1-20) as well as the dispersion concentration (0.1 to 5 mg/mL) using the vacuum filtration method. In the in-plane, four different electrical responses were observed and modeled: (1) a substrate dominated spectrum representing unconnected MWNTs, (2) one that included bundle and junction responses as well as some inductances representing sparsely distributed MWNT networks, (3) followed by a parallel RL circuit for partially connected MWNT networks and (4) finally a series RL circuit for fully connected MWNT networks. In the thru-plane, only two different electrical responses were observed and modeled. The results for the in- plane and thru-plane properties were used to generate percolation curves that show that electrical conductivity can change as much as 10 orders of magnitude for the same exact MWNTs. Similar experiments were also carried out for films that were spin coated and inkjet printed. The specific nanotube arrangements and electrical response depend on the number of layers deposited but in principle the trends observed are the same given that we used the SAME starting nanotube materials. Our results indicate that not only do the characteristics of the nanotubes themselves play a role but also the structure of the underlying substrate and the details of how the films are deposited.