Because of their small size and large specific surface area (SA), insoluble nanoparticles are almost not affected by the gravitational force and are generally formulated in stable suspensions or sols. This raises, however, a potential difficulty in in vitro assay systems in which cells adhering to the bottom of a culture vessel may not be exposed to the majority of nanoparticles in suspension. J. G. Teeguarden et al., 2007, Toxicol. Sci. 95, 300-312 have recently addressed this issue theoretically, emphasizing the need to characterize the effective dose (mass or number or SA dose of particles that affect the cells) which, according to their model based on sedimentation and gravitation forces, might only represent a very small fraction of the nominal dose. We hypothesized, in contrast, that because of convection forces that usually develop in sols, the majority of the particles may reach the target cells and exert their potential toxicity. To address this issue, we exposed three different cell lines (A549 epithelial cells, EAHY926 endothelial cells, and J774 monocyte-macrophages) to a monodisperse suspension of Stöber silica nanoparticles (SNP) in three different laboratories. Four different end points (lacticodehydrogenase [LDH] release, LDH cell content, tetrazolium salt (MTT), and crystal violet staining) were used to assess the cell response to nanoparticles. We found, in all cell lines and for all end points, that the cellular response was determined by the total mass/number/SA of particles as well as their concentration. Practically, for a given volume of dispersion, both parameters are of course intimately interdependent. We conclude that the nominal dose remains the most appropriate metric for in vitro toxicity testing of insoluble SNP dispersed in aqueous medium. This observation has important bearings on the experimental design and the interpretation of in vitro toxicological studies with nanoparticles.