Magnetic nanoparticles [MNPs] made from iron oxides have many applications in biomedicine. Full understanding of the interactions between MNPs and mammalian cells is a critical issue for their applications. In this study, MNPs were coated with poly(ethylenimine) [MNP-PEI] and poly(ethylene glycol) [MNP-PEI-PEG] to provide a subtle difference in their surface charge and their cytotoxicity which were analysed by three standard cell viability assays: 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium [MTS], CellTiter-Blue and CellTiter-Glo (Promega, Southampton, UK) in SH-SY5Y and RAW 264.7 cells The data were validated by traditional trypan blue exclusion. In comparison to trypan blue manual counting, the MTS and Titer-Blue assays appeared to have consistently overestimated the viability. The Titer-Glo also experienced a small overestimation. We hypothesise that interactions were occurring between the assay systems and the nanoparticles, resulting in incorrect cell viability evaluation. To further understand the cytotoxic effect of the nanoparticles on these cells, reactive oxygen species production, lipid peroxidation and cell membrane integrity were investigated. After pegylation, the MNP-PEI-PEG possessed a lower positive surface charge and exhibited much improved biocompatibility compared to MNP-PEI, as demonstrated not only by a higher cell viability, but also by a markedly reduced oxidative stress and cell membrane damage. These findings highlight the importance of assay selection and of dissection of different cellular responses in in-vitro characterisation of nanostructures.