Highly conductive polymer nanocomposites, incorporating nanostructures based on carbon allotropes including 3-D graphite, 2-D graphene and its derivatives, 1-D carbon nanotubes and 0-D fullerenes, are interesting materials for novel, cost-effective high-performance applications. Graphite nanoplates (GnP) consist of ultrathin stacks of graphene layers, having high stiffness (1 TPa) and good electrical (~106 Ω.cm-1) and in-plane thermal (~2000 W.m-1.K-1) conductivity. They are usually obtained through intercalation and exfoliation of natural flaky graphite, and thus are inexpensive when compared with other carbon fillers. However, in practice, the effective dispersion of GnP into polymeric matrices remains a critical issue, due to their inherent tendency to form clusters with strong cohesive strength and lack of chemical functionalities at the surface and edges, which precludes the possibility to establish strong interfaces with the matrix. When preparing nanocomposites by melt mixing, namely using conventional polymer compounding technologies, dispersion is influenced by material properties (aspect ratio, surface chemistry, purity, agglomerate density and strength of the filler) and processing conditions (temperature, screw speed, output and screw profile). Also, re-agglomeration of GnP can developed when the thermomechanical stresses are relieved. This work focus on the effects of morphology and surface chemistry of GnP on the kinetics of dispersion, re-agglomeration and network formation (with emphasis on the resulting electrical conductivity). The challenges to ensure adequate dispersion are discussed.