Creation of high-quality and novel polymer particle nanocomposites to a large extent depends on understanding the behaviors of individual polymer chains and particles, especially at the mixing state in a liquid solvent. Simulations can help identify critical parameters and equations that govern the suspension behaviors. This study is the first attempt to understand the agglomeration processes of ZnO nanoparticle and poly(methyl methacrylate) polymer cosuspensions through a constant number Monte Carlo simulation. A modified Derjaguin-Landau-Verwey-Overbeek theory is used to describe the particle-particle interactions that lead to agglomeration. The average agglomerate size and number are measured as a function of suspension resting time, particle to polymer volume ratio, polymer chain length, and suspension drying. The agglomerate size increases persistently with the resting time and particle content increase, ranging from 1.2 mu m for the 1 vol % particle content suspension to 4.6 mu m for the 20 vol % particle content suspension after 30 min of suspension resting. The agglomerate size distribution for all of the particle contents follows a lognormal distribution. As the polymer chain length increases, agglomeration also becomes more severe. If drying is accounted for and thus the solids loading continually increases, the suspension becomes much more stable because of increases in viscosity and depletion stabilization.