The constancy of cell volume under physiological conditions is generally thought to reflect a balance between solute influx and efflux and is therefore critically dependent on the properties of the plasma membrane. Despite a number of unanswered questions, a large amount of experimental data can be accommodated within this pump-leak framework and can by analysed using the simple assumptions of osmotic equality between cells and their surrounding fluid, and fluid electroneutrality. Experimentally, cell volume may be altered in vitro either by changing cell solute content under isosmotic conditions or by changing medium osmolality. Exposure to anisosmotic media may provoke a variety of cell responses that minimise the volume change. However, much of this experimental work has been performed under extreme conditions in vitro that would never be experienced by vertebrate cells in vivo; its relevance to pathophysiological situations is questionable. It is argued that regulation of cell volume should not be seen in isolation but as part of the process, cell homeostasis, by which cells attempt to minimise changes in composition when faced with perturbations in their environment. Given the variety of processes and the large numbers of membrane transporters, an understanding of how cells respond to such perturbations requires a combination of modelling and experimentation. A simple example of this approach is presented.