The aim of this study is to optimize the sample preparation protocol for dispersing powder-form ceramic nanoparticles (CeNPs) in liquid media by leveraging the power of the design of experiment approach to narrow down potential causes of aggregation and to investigate the subsequent effect of CeNP aggregation on the biological activity of SaOS-2 human osteosarcoma cells. Dispersion of the nanopowders is one of the problems in the nanotechnology field because of the tendency of the nanoparticles for aggregation. Although there are existing dispersion protocols, they offer a one-size-fits-all approach overseeing the unique physicochemical properties of the different nanomaterials. In this study, optimization of the sample preparation protocol for two CeNPs was assessed via the investigation of the most contributing parameters and their synergetic effect through measurements of Z-average and zeta potential. Evaluation of these parameters allowed the development of two different models for each nanomaterial, predicting Z-average and zeta potential for given parameter sets. Through these models, two different sample sets were selected to evaluate the effect of aggregation on the SaOS-2 cell line. Outcomes show that the concentration of nanomaterial, pH, and the presence of additive molecule are three main parameters that affect dispersion stability. It was seen that these parameters can be included in a design to develop an efficient model to predict Z-average and zeta potential for investigated nanomaterials. Moreover, cell viability tests show that there is no significant difference between untreated and nanomaterial-treated cells. The findings promise that tailor-made and reliable dispersion protocols for different nanopowders can be developed via design of experiment.