New insulation materials can be tailored by compounding nanofillers and polymer materials. The advantageous changes in properties of polymer nanocomposites can be explained by the large interface area of the nanofillers. Theoretical models developed in the field of nanodielectrics concentrate on the analysis of the interface volume. Dielectric breakdown strength (DBS) is one of the most important properties of an insulation material. The DBS of nanocomposites depends heavily on the nanofiller content and even small quantities can cause improvement. This is linked to the maximum interface volume achieved even at low nanofiller concentrations. Homogeneous nanodispersion is the key for DBS increase and reliable results. Often the most profitable nanofiller quantities are below 5 wt-%. A similar observation was made with relative permittivity and dielectric losses of nanocomposites. At low nanofiller concentrations the relative permittivity and dielectric losses stay at the reference level or are even lower. With increasing nanofiller content relative permittivity and dielectric losses start to increase. This is related to the overlapping of interfacial zones of nanoparticles. The overlapping of the interfacial zones also depends on the nanofiller size and dispersion. The treeing growth and partial discharge (PD) endurance of nanocomposites was found to have a partly different kind of behaviour. Surface PD endurance of nanocomposites is strongly a mass related phenomenon, but treeing growth is more complex. It is related to large interface volume of nanocomposites having similar behaviour as observed with short-time DBS. In this dissertation the above mentioned dielectric properties are studied experimentally, concentrating on silica (SiO2)-polypropylene (PP), polyhedral oligomeric silsesquixane (POSS)-PP and POSS-epoxy (EP) nanocomposites. The dielectric properties are verified for sheet samples. In DBS measurements ac, dc and lightning impulse (LI) voltage shapes are used. Relative permittivity and dielectric losses are measured with insulation diagnosis analyzer (IDA) or with arbitrary waveform impedance spectroscopy (AWIS). PD endurance of nanocomposites is investigated with rounded rod- or conical rod-plane electrode setups. Proper nanodispersion is reflected in improved dielectric behaviour of 5 wt-% SiO2-PP nanocomposite. 5 wt-% SiO2-PP nanocomposite showed a promising combination of dielectric properties especially considering ac and dc power capacitors. The ac and dc breakdown strength of 5 wt-% SiO2-PP nanocomposite increased 20 % and 52 % respectively. It is equally important that with both ac and dc voltages the β value in the Weibull probability function is higher with the nanocomposite than with the reference PP, indicating a smaller standard deviation and more homogeneous material. These results have an interesting impact on insulation system design, because safety margins and insulation thicknesses could be reduced. With 5 wt-% SiO2-PP nanocomposite dielectric losses remained comparable to reference PP. Capacitance did decrease with both materials until 80 °C. This is important, considering e.g. power capacitors, which operate in the 50-60 °C temperature region. There is no capacitance change compared to reference PP and no additional heat is produced due to low loss factor. In PD endurance measurements the average lifetime of 5 wt-% SiO2-PP nanocomposite was 9 times longer than with reference PP. Also, at the same voltages the PD levels were higher for reference PP than for 5 wt-% SiO2-PP nanocomposite. Surface analysis of the films revealed that a layer of silicon oxides formed on the surfaces of the 5 wt-% SiO2-PP nanocomposite during PD stress. The layers may protect this nanocomposite from degradation and lengthen the lifetime of the sample. This shielding behaviour is significant because the layers are formed due to PD stress to critical spots of insulation. With respect to the properties of nanodielectrics studied possible applications could be e.g. capacitors, cables, dry transformers, rotating machines, switchgear and outdoor insulations. Although the future of nanodielectrics is promising, there are many questions to be resolved. Manufacturing processes must be optimized and up-scaled to larger volumes. Promising results have mainly been achieved with short-term measurements and work has been done to understand mechanisms behind the functionality, which is important. Full scale models of applications should be produced and long-term measurements conducted in operating conditions. Long-term measurements for real components will finally reveal the actual benefits and possible drawbacks of nanocomposites compared to the insulation materials used at the moment. In addition to dielectric properties, the thermal, mechanical, chemical and ageing properties critical for each component must be verified. This means that the industry should be intensively linked to develop actual new products.