Short-term, within one 24-hour day (diurnal period) within-person changes of the principal blood count analytes in healthy subjects were studied at three major institutions. The results from each test site were indistinguishable and were therefore combined to make a database of 96 healthy subjects. Analytical imprecision of each analyte was subtracted from the total observed variation to give true diurnal change. Each analyte showed characteristic changes. As would be expected, cellular properties of erythrocytes, such as MCV (mean cell volume) and MCH (mean cell hemoglobin) showed negligible change. The red cell count, hematocrit, and hemoglobin showed changes that were consistent with fluid balance change. Total white cell count and some differential count components showed major changes that raised questions of the confidence limits of clinical decision levels and the validity of commonly used reference intervals. Platelet count changes were typically less than analytic imprecision, suggesting the need for improvement in this aspect of analyzer performance.