Dispersion and agglomeration control the macroscopic properties of nanocomposites; thus, quantitative characterization of particle dispersion and agglomeration is crucial. So far, methods such as image analysis or measurement of macroscopic properties have been insufficient. So, a novel methodology is introduced to predict particle dispersion and agglomeration degree quantitatively through the measurement of effective dielectric constant. This is done by taking into account the impact of inter-particle interactions in respect to microstructure for the calculation of effective dielectric constant. Inter-particle interaction is calculated as a function of inter-particle distance for random dispersion of particles in composites and as a function of agglomerate size and packing density for nanocomposites with agglomerated microstructure. The inter-particle interactions increase with increase in filler content and agglomeration. This increment is abrupt at percolation where inter-particle distance is significantly reduced. Experimental data from ZnO/PS composites with controlled inter-particle distance and agglomeration confirm the theoretical results.