Polymeric materials containing nanometer (nm) size particles are being introduced to provide compact shapes for low and medium voltage insulation equipment. The nanocomposites may provide superior electrical performance when compared with those available currently, such as lower dielectric losses and increased dielectric strength, tracking and erosion resistance, and surface hydrophobicity. All of the above mentioned benefits can be achieved at a lower filler concentration (< 10%) than conventional microfillers (40-60%). Also, the uniform shapes of nanofillers provide a better electrical stress distribution as compared to irregular shaped microcomposites which can have high internal electric stress, which could be a problem for devices with active electrical parts. Improvement in electrical performance due to addition of nanofillers in an epoxy matrix has been evaluated in this work. Scanning Electron Microscopy (SEM) was done on the epoxy samples to confirm uniform dispersion of nano-sized fillers as good filler dispersion is essential to realize the above stated benefits. Dielectric spectroscopy experiments were conducted over a wide range of frequencies as a function of temperature to understand the role of space charge and interfaces in these materials. The experiment results demonstrate significant reduction in dielectric losses in samples containing nanofillers. High voltage experiments such as corona resistance tests were conducted over 500 hours to monitor degradation in the samples due to corona. These tests revealed improvements in partial discharge endurance of nanocomposite samples. These improvements could not be adequately explained using a macroscopic quantity such as thermal conductivity. Thermo gravimetric analysis (TGA) showed higher weight loss initiation temperatures for nanofilled samples which is in agreement with the corona resistance experimental results. Theoretical models have also been developed in this work to complement the results of the corona resistance experiment and the TGA analysis. Degradation model was developed to map the erosion path using Dijkstra's shortest path algorithm. A thermal model was developed to calculate the localized temperature distribution in the micro and nano-filled samples using the PDE toolbox in MATLAB. Both the models highlight the fact that improvement in nanocomposites is not limited to the filler concentrations that were tested experimentally.