A method for the assessment of red cell filterability with paper filters is described. Two milliliters of a washed red cell suspension with a hematocrit of 50% was filtered through a filter paper cone in a glass funnel and the filtration half-time (FT1/2) was measured. The filter papers were calibrated by previous measurement of the filtration time for the suspending solution, the calibration time (CT). A linear correlation between CT and FT1/2 was found for individual red cell suspensions (p less than 0.001) as well as for a normal population (p less than 0.001). Consequently, the index FT1/2/CT should represent a more reproducible parameter than FT1/2 and experimentally this appears to be the case. In fact, elimination of erratic values in a normal population was realized. Furthermore, discrimination between normal and abnormal values was improved and the coefficient of variation of single measurements passed from 28.4 to 9.4%. The reproducibility and reliability of this extremely simple method are therefore decisively enhanced without complication of the procedure.