A protocol for the statistical and graphical evaluation of erythrocyte volume distributions has been developed using expectation-maximization methods for distribution parameter estimation. In a reference sample group of 50 healthy individuals, none of the red cell volume distributions determined by an aperture impedance method differed significantly from a log normal model. Measurements in 211 patients with anemia provided evidence that similar statistical methods could be used for the characterization of abnormal red cell volume distributions. Graphical displays accurately summarize the quantitative information derived from the analysis of volume distributions and show a reference distribution fitted to the observed distribution against the range of distributions found in healthy individuals. This protocol provides a specific, statistically valid implementation of the general principles for the analysis of cell volume distributions proposed by the International Committee for Standardization in Hematology.