We evaluated whether any monosaccharides inhibit glycolysis in erythrocytes and discovered that D-mannose does. In the presence of D-mannose, glucose can be accurately measured by either the hexokinase procedure or the glucose oxidase procedure. In comparison studies with other glucose preservatives, we found that after 2 h at room temperature glucose decreased by 21 (SD 13) mg/L in D-mannose-treated blood, 93 (SD 10) mg/L in sodium fluoride-treated blood, 28 (SD 21) mg/L in ice-cooled blood, and 144 (SD 28) mg/L in control blood (no preservative treatment). Because D-mannose acted in the early phase of glycolysis, it was a more effective preservative than sodium fluoride; moreover, its use did not preclude measurement of sodium and potassium in the blood samples. D-Mannose did not interfere with other routine chemical tests except for the assay of creatine kinase involving coupled enzymes hexokinase/glucose-6-phosphate dehydrogenase. Creatine kinase could be correctly assayed in the presence of D-mannose by using glucokinase instead of hexokinase. D-Mannose can be used with or without anticoagulant and is compatible with most types of multi-channel automated analyzers.