The quality and performance of nanocomposites is usually associated with the extent of dispersion of the nanoparticles in the matrix. Perfect dispersion is normally characterized, at the ideal limit, as the achievement of a uniform distribution of exfoliated (individual) nanoparticle within the system. In fact, much of the published work on nanocomposites seems to deal with structures involving nanoparticle aggregates of different sizes with enhanced separation (e.g. d-spacing) or disentanglement. In many instances, some large agglomerates also appear in the product. Various physical and chemical factors contribute to this behavior. Some of the important factors relate to either the processing conditions or to the state of the structure of the initial particles employed in the preparation of the nanocomposite. The presentation will discuss some of the above issues and illustrate the role of these factors, especially the initial structure of the particle agglomerates, in the determination of the final state of dispersion. Emphasis will be placed on experimental observations obtained with systems incorporating clay and nanocrystalline cellulose (NCC) as the sources of nanoparticles.