A careful analysis of platelet counting methods, including an assessment of accuracy, was done using data from the College of American Pathologists Comprehensive Hematology Survey preserved platelet suspension specimen (H-32-1976). Comparisons between methods suggest systematic biases probably related to the calibration methods used for automated instruments. Falsely elevated counts occurred with manual methods and one light-dispersion system. Methods for calibration and continuing quality assurance of platelet counting are proposed.
