Cumulative methods of particle size analysis are sometimes alleged to be inaccurate because it is necessary to differentiate the sedimentation curve in order to evaluate the size distribution. By considering a hypothetical distribution it has been shown that, provided sufficient points are obtained to enable the sedimentation curve to be plotted accurately, this differentiation does not give rise to errors. Comparative tests with the microscope and sedimentation balance on two samples composed of spherical particles confirm this conclusion.