Pneumotachometer is the most widely used respiratory air flow sensor. But the measurement error hard to be ignored exists, since the empirical calibration carve in a closed form cannot provide accurate flow values at all measured pressure data. Any possible distortion of the calibration curve further increases measurement error, which cannot he corrected by the conventional calibration method. The present study suggests a new error correction algorithm applicable to any type of pneumotachometer equipped with computer-based data processing device. This algorithm evaluates conductance error correction coefficients at each digitized pressure level based on the cumulative pressure data obtained by manual volume syrange stroke operation at different speeds. Experiment demonstrated error minimization down to below 1/10 in mean stroke volume measurement compared to the conventional calibration curve method. The standard deviation of the stroke volume also decreased down to below 1/2 of the conventional method. The present technique can be introduced with no difficulty to the calibration procedure the American Thoracic Society, and expected to guarantee the accuracy and reliability of the respiratory air flow measurement during spirometric tests.