In a chronoamperometric measurement the chemical activity of glucose is directly determined using a glucose sensitive biosensor based on glucose oxidase and an efficient mediator. Within the physiological range of 2 mmol kg À1 and 10 mmol kg À1 glucose in aqueous solutions behaves almost ideally and the chemical activity equals the glucose molality, the amount of glucose per unit mass of water, within the estimated measurement uncertainty. In physiological samples the equality between chemical activity and glucose molality is destroyed and has to be corrected for. Gravimetrically prepared mixtures of high purity glucose in buffered aqueous solutions are used as standards for calibration of the chronoamperometric measurement setup. Based on the model equation for the measurand aqueous or physiological sample, all significant sources of uncertainty are identified, their magnitude estimated from published and experimental data and finally combined to give the uncertainty in the reported value of the glucose molality. It is found, that the combined uncertainty of the glucose molality comprises mainly uncertainty contributions from the nonideal behavior, the chronoamperometric measurement setup, from the purity of glucose used and from the chemical composition of the test sample. The expanded uncertainty is below 2% rel., the glucose content determined by the bioelectrochemical measurements thus competes well with todays considered most accurate reference method Isotope Dilution Mass Spectrometry. Advantageous of the presented electroanalytical method is the direct measure of the glucose molality without prior sample preparation and dilution.