Cell division and apoptosis are normal processes that are necessary for human health when kept in proper balance. Upsetting this balance can lead to a variety of diseases. Change in cell volume is a fundamental and probably essential feature of both cell division and apoptosis, so it is important to understand mechanisms of cell volume control. Particularly in the case of apoptosis, this mechanism is not well understood. Studying cell volume control requires an efficient method of measuring cell size in populations of cells. The method of choice is usually flow cytometry, which involves expensive equipment and high running costs that may tend to prohibit speculative pilot studies. We aimed to develop a cheap and easy method for measuring cell size that could generate pilot data to be followed up with more sophisticated methods. CellProfiler (1) is a software package that was designed for the automated analysis of fluorescence images, which feature high contrast between labelled cells and the background, enabling the software to identify cells in the image. On the other hand, phase contrast images tend not to have a high contrast between the cell interior and the background, but show an edge at the boundary of the cell. CellProfiler has an edge finder module that we have exploited in developing a pipeline that can automatically measure the sizes of cells in a large number of phase contrast images. The pipeline also incorporates upper and lower size cutoffs to rule out cell clusters and debris. Phase contrast images were taken of Jurkat cells in culture with a Nikon Diaphot inverted microscope fitted with a Nikon D-40 digital camera. Images were saved as jpg files for analysis by CellProfiler. The pipeline was validated against manual analysis using Scion Image software (Scion Corporation, Frederick, MD, USA). The jpg files were converted to tif files for Scion Image. In 5 images of well separated Jurkat cells in culture, manual analysis identified 236 cells with profile areas of (mean ± s.d.) 6608 ± 2466 pixels, while CellProfiler identified 185 cells with profile areas of 6190 ± 3175 pixels. Although CellProfiler missed some cells, this result indicates that it could be a useful tool for analysing changes in the size distribution of cell populations imaged with phase-contrast microscopy.